diff --git a/.eslintignore b/.eslintignore
index 53cfa65d36c6..d25c116b8537 100644
--- a/.eslintignore
+++ b/.eslintignore
@@ -4,6 +4,7 @@ bin/
 flow-typed/**
 packages/*/build/**
 packages/*/build-es5/**
+packages/jest-matcher-utils/src/cleanupSemantic.ts
 website/blog
 website/build
 website/node_modules
diff --git a/.prettierignore b/.prettierignore
index 1019f4c475e3..9b259cd268a9 100644
--- a/.prettierignore
+++ b/.prettierignore
@@ -1,3 +1,4 @@
 fixtures/failing-jsons/
+packages/jest-matcher-utils/src/cleanupSemantic.ts
 packages/jest-config/src/__tests__/jest-preset.json
 packages/pretty-format/perf/world.geo.json
diff --git a/CHANGELOG.md b/CHANGELOG.md
index ea66aee969ad..13bd95c916d4 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -2,6 +2,7 @@
 
 ### Features
 
+- `[expect]` Highlight substring differences when matcher fails, part 1 ([#8448](https://github.com/facebook/jest/pull/8448))
 - `[jest-cli]` Improve chai support (with detailed output, to match jest exceptions) ([#8454](https://github.com/facebook/jest/pull/8454))
 
 ### Fixes
diff --git a/packages/expect/src/__tests__/__snapshots__/matchers.test.js.snap b/packages/expect/src/__tests__/__snapshots__/matchers.test.js.snap
index f09ede5af303..2de490df82a3 100644
--- a/packages/expect/src/__tests__/__snapshots__/matchers.test.js.snap
+++ b/packages/expect/src/__tests__/__snapshots__/matchers.test.js.snap
@@ -278,6 +278,13 @@ exports[`.toBe() fails for 'undefined' with '.not' 1`] = `
 Expected: not <green>undefined</>"
 `;
 
+exports[`.toBe() fails for: "" and "compare one-line string to empty string" 1`] = `
+"<dim>expect(</><red>received</><dim>).</>toBe<dim>(</><green>expected</><dim>) // Object.is equality</>
+
+Expected: <green>\\"compare one-line string to empty string\\"</>
+Received: <red>\\"\\"</>"
+`;
+
 exports[`.toBe() fails for: "abc" and "cde" 1`] = `
 "<dim>expect(</><red>received</><dim>).</>toBe<dim>(</><green>expected</><dim>) // Object.is equality</>
 
@@ -303,6 +310,13 @@ string" 1`] = `
 <dim>  string</>"
 `;
 
+exports[`.toBe() fails for: "painless JavaScript testing" and "delightful JavaScript testing" 1`] = `
+"<dim>expect(</><red>received</><dim>).</>toBe<dim>(</><green>expected</><dim>) // Object.is equality</>
+
+Expected: <green>\\"<inverse>delightful</> JavaScript testing\\"</>
+Received: <red>\\"<inverse>painless</> JavaScript testing\\"</>"
+`;
+
 exports[`.toBe() fails for: "with 
 trailing space" and "without trailing space" 1`] = `
 "<dim>expect(</><red>received</><dim>).</>toBe<dim>(</><green>expected</><dim>) // Object.is equality</>
@@ -1922,6 +1936,13 @@ exports[`.toContain(), .toContainEqual() error cases for toContainEqual 1`] = `
 Received has value: <red>null</>"
 `;
 
+exports[`.toEqual() {pass: false} expect("1 234,57 $").toEqual("1 234,57 $") 1`] = `
+"<dim>expect(</><red>received</><dim>).</>toEqual<dim>(</><green>expected</><dim>) // deep equality</>
+
+Expected: <green>\\"1<inverse> </>234,57<inverse> </>$\\"</>
+Received: <red>\\"1<inverse> </>234,57<inverse> </>$\\"</>"
+`;
+
 exports[`.toEqual() {pass: false} expect("Eve").toEqual({"asymmetricMatch": [Function asymmetricMatch]}) 1`] = `
 "<dim>expect(</><red>received</><dim>).</>toEqual<dim>(</><green>expected</><dim>) // deep equality</>
 
@@ -1985,15 +2006,8 @@ exports[`.toEqual() {pass: false} expect([1]).toEqual([2]) 1`] = `
 exports[`.toEqual() {pass: false} expect({"a": 1, "b": 2}).toEqual(ObjectContaining {"a": 2}) 1`] = `
 "<dim>expect(</><red>received</><dim>).</>toEqual<dim>(</><green>expected</><dim>) // deep equality</>
 
-<green>- Expected</>
-<red>+ Received</>
-
-<green>- ObjectContaining {</>
-<green>-   \\"a\\": 2,</>
-<red>+ Object {</>
-<red>+   \\"a\\": 1,</>
-<red>+   \\"b\\": 2,</>
-<dim>  }</>"
+Expected: <green>ObjectContaining {\\"a\\": 2}</>
+Received: <red>{\\"a\\": 1, \\"b\\": 2}</>"
 `;
 
 exports[`.toEqual() {pass: false} expect({"a": 1}).toEqual({"a": 2}) 1`] = `
@@ -3057,6 +3071,15 @@ Expected value: <green>2</>
 Received value: <red>1</>"
 `;
 
+exports[`.toHaveProperty() {pass: false} expect({"children": ["\\"That cartoon\\""], "props": null, "type": "p"}).toHaveProperty('children,0', "\\"That cat cartoon\\"") 1`] = `
+"<dim>expect(</><red>received</><dim>).</>toHaveProperty<dim>(</><green>path</><dim>, </><green>value</><dim>)</>
+
+Expected path: <green>[\\"children\\", 0]</>
+
+Expected value: <green>\\"\\\\\\"That <inverse>cat </>cartoon\\\\\\"\\"</>
+Received value: <red>\\"\\\\\\"That cartoon\\\\\\"\\"</>"
+`;
+
 exports[`.toHaveProperty() {pass: false} expect({"key": 1}).toHaveProperty('not') 1`] = `
 "<dim>expect(</><red>received</><dim>).</>toHaveProperty<dim>(</><green>path</><dim>)</>
 
@@ -3497,6 +3520,13 @@ Expected substring: <green>\\"foo\\"</>
 Received string:    <red>\\"bar\\"</>"
 `;
 
+exports[`.toStrictEqual() displays substring diff 1`] = `
+"<dim>expect(</><red>received</><dim>).</>toStrictEqual<dim>(</><green>expected</><dim>) // deep equality</>
+
+Expected: <green>\\"<inverse>Another caveat is that</> Jest will not typecheck your tests.\\"</>
+Received: <red>\\"<inverse>Because TypeScript support in Babel is just transpilation,</> Jest will not type<inverse>-</>check your tests<inverse> as they run</>.\\"</>"
+`;
+
 exports[`.toStrictEqual() matches the expected snapshot when it fails 1`] = `
 "<dim>expect(</><red>received</><dim>).</>toStrictEqual<dim>(</><green>expected</><dim>) // deep equality</>
 
diff --git a/packages/expect/src/__tests__/matchers.test.js b/packages/expect/src/__tests__/matchers.test.js
index 86daa26117ae..de742ae50bcf 100644
--- a/packages/expect/src/__tests__/matchers.test.js
+++ b/packages/expect/src/__tests__/matchers.test.js
@@ -215,6 +215,8 @@ describe('.toBe()', () => {
     [Symbol('received'), Symbol('expected')],
     [new Error('received'), new Error('expected')],
     ['abc', 'cde'],
+    ['painless JavaScript testing', 'delightful JavaScript testing'],
+    ['', 'compare one-line string to empty string'],
     ['with \ntrailing space', 'without trailing space'],
     ['four\n4\nline\nstring', '3\nline\nstring'],
     [[], []],
@@ -318,6 +320,16 @@ describe('.toStrictEqual()', () => {
     ).toThrowErrorMatchingSnapshot();
   });
 
+  it('displays substring diff', () => {
+    const expected =
+      'Another caveat is that Jest will not typecheck your tests.';
+    const received =
+      'Because TypeScript support in Babel is just transpilation, Jest will not type-check your tests as they run.';
+    expect(() =>
+      jestExpect(received).toStrictEqual(expected),
+    ).toThrowErrorMatchingSnapshot();
+  });
+
   it('does not pass for different types', () => {
     expect({
       test: new TestClassA(1, 2),
@@ -358,6 +370,7 @@ describe('.toEqual()', () => {
     [{a: 1}, {a: 2}],
     [{a: 5}, {b: 6}],
     ['banana', 'apple'],
+    ['1\u{00A0}234,57\u{00A0}$', '1 234,57 $'], // issues/6881
     [null, undefined],
     [[1], [2]],
     [[1, 2], [2, 1]],
@@ -1348,6 +1361,14 @@ describe('.toHaveProperty()', () => {
   const memoized = function() {};
   memoized.memo = [];
 
+  const receivedDiff = {
+    children: ['"That cartoon"'],
+    props: null,
+    type: 'p',
+  };
+  const pathDiff = ['children', 0];
+  const valueDiff = '"That cat cartoon"';
+
   [
     [{a: {b: {c: {d: 1}}}}, 'a.b.c.d', 1],
     [{a: {b: {c: {d: 1}}}}, ['a', 'b', 'c', 'd'], 1],
@@ -1384,6 +1405,7 @@ describe('.toHaveProperty()', () => {
     [{a: {b: {c: {d: 1}}}}, 'a.b.c.d', 2],
     [{'a.b.c.d': 1}, 'a.b.c.d', 2],
     [{'a.b.c.d': 1}, ['a.b.c.d'], 2],
+    [receivedDiff, pathDiff, valueDiff],
     [{a: {b: {c: {d: 1}}}}, ['a', 'b', 'c', 'd'], 2],
     [{a: {b: {c: {}}}}, 'a.b.c.d', 1],
     [{a: 1}, 'a.b.c.d', 5],
diff --git a/packages/expect/src/matchers.ts b/packages/expect/src/matchers.ts
index 977cb205cbb3..52a8c9fe7e4f 100644
--- a/packages/expect/src/matchers.ts
+++ b/packages/expect/src/matchers.ts
@@ -18,6 +18,7 @@ import {
   getLabelPrinter,
   matcherErrorMessage,
   matcherHint,
+  printDiffOrStringify,
   printReceived,
   printExpected,
   printWithType,
@@ -26,7 +27,6 @@ import {
 } from 'jest-matcher-utils';
 import {MatchersObject, MatcherState} from './types';
 import {
-  printDiffOrStringify,
   printExpectedConstructorName,
   printExpectedConstructorNameNot,
   printReceivedArrayContainExpectedItem,
@@ -51,6 +51,9 @@ const RECEIVED_LABEL = 'Received';
 const EXPECTED_VALUE_LABEL = 'Expected value';
 const RECEIVED_VALUE_LABEL = 'Received value';
 
+// The optional property of matcher context is true if undefined.
+const isExpand = (expand?: boolean): boolean => expand !== false;
+
 const toStrictEqualTesters = [
   iterableEquality,
   typeEquality,
@@ -107,7 +110,7 @@ const matchers: MatchersObject = {
               received,
               EXPECTED_LABEL,
               RECEIVED_LABEL,
-              this.expand,
+              isExpand(this.expand),
             )
           );
         };
@@ -577,7 +580,7 @@ const matchers: MatchersObject = {
             received,
             EXPECTED_LABEL,
             RECEIVED_LABEL,
-            this.expand,
+            isExpand(this.expand),
           );
 
     // Passing the actual and expected objects so that a custom reporter
@@ -751,7 +754,7 @@ const matchers: MatchersObject = {
                 receivedValue,
                 EXPECTED_VALUE_LABEL,
                 RECEIVED_VALUE_LABEL,
-                this.expand,
+                isExpand(this.expand),
               )
             : `Received path: ${printReceived(
                 expectedPathType === 'array' || receivedPath.length === 0
@@ -886,7 +889,7 @@ const matchers: MatchersObject = {
             getObjectSubset(received, expected),
             EXPECTED_LABEL,
             RECEIVED_LABEL,
-            this.expand,
+            isExpand(this.expand),
           );
 
     return {message, pass};
@@ -918,7 +921,7 @@ const matchers: MatchersObject = {
             received,
             EXPECTED_LABEL,
             RECEIVED_LABEL,
-            this.expand,
+            isExpand(this.expand),
           );
 
     // Passing the actual and expected objects so that a custom reporter
diff --git a/packages/expect/src/print.ts b/packages/expect/src/print.ts
index c386d388b769..aa03601398d8 100644
--- a/packages/expect/src/print.ts
+++ b/packages/expect/src/print.ts
@@ -6,18 +6,13 @@
  *
  */
 
-import getType, {isPrimitive} from 'jest-get-type';
 import {
   EXPECTED_COLOR,
   INVERTED_COLOR,
   RECEIVED_COLOR,
-  diff,
-  getLabelPrinter,
-  printExpected,
   printReceived,
   stringify,
 } from 'jest-matcher-utils';
-import {isOneline} from './utils';
 
 // Format substring but do not enclose in double quote marks.
 // The replacement is compatible with pretty-format package.
@@ -66,74 +61,6 @@ export const printReceivedArrayContainExpectedItem = (
       ']',
   );
 
-const shouldPrintDiff = (expected: unknown, received: unknown): boolean => {
-  const expectedType = getType(expected);
-  const receivedType = getType(received);
-
-  if (expectedType !== receivedType) {
-    return false;
-  }
-
-  if (isPrimitive(expected)) {
-    // Print diff only if both strings have more than one line.
-    return expectedType === 'string' && !isOneline(expected, received);
-  }
-
-  if (
-    expectedType === 'date' ||
-    expectedType === 'function' ||
-    expectedType === 'regexp'
-  ) {
-    return false;
-  }
-
-  if (expected instanceof Error && received instanceof Error) {
-    return false;
-  }
-
-  return true;
-};
-
-export const printDiffOrStringify = (
-  expected: unknown,
-  received: unknown,
-  expectedLabel: string, // include colon and one or more spaces,
-  receivedLabel: string, // same as returned by getLabelPrinter
-  expand?: boolean, // diff option: true if `--expand` CLI option
-): string => {
-  // Cannot use same serialization as shortcut to avoid diff,
-  // because stringify (that is, pretty-format with min option)
-  // omits constructor name for array or object, too bad so sad :(
-  const difference = shouldPrintDiff(expected, received)
-    ? diff(expected, received, {
-        aAnnotation: expectedLabel,
-        bAnnotation: receivedLabel,
-        expand,
-      }) // string | null
-    : null;
-
-  // Cannot reuse value of stringify(received) in report string,
-  // because printReceived does inverse highlight space at end of line,
-  // but RECEIVED_COLOR does not (it refers to a plain chalk method).
-  if (
-    typeof difference === 'string' &&
-    difference.includes('- ' + expectedLabel) &&
-    difference.includes('+ ' + receivedLabel)
-  ) {
-    return difference;
-  }
-
-  const printLabel = getLabelPrinter(expectedLabel, receivedLabel);
-  return (
-    `${printLabel(expectedLabel)}${printExpected(expected)}\n` +
-    `${printLabel(receivedLabel)}${
-      stringify(expected) === stringify(received)
-        ? 'serializes to the same string'
-        : printReceived(received)
-    }`
-  );
-};
-
 export const printExpectedConstructorName = (
   label: string,
   expected: Function,
diff --git a/packages/jest-matcher-utils/package.json b/packages/jest-matcher-utils/package.json
index 59e468f4632d..66d758341438 100644
--- a/packages/jest-matcher-utils/package.json
+++ b/packages/jest-matcher-utils/package.json
@@ -14,6 +14,7 @@
   "main": "build/index.js",
   "dependencies": {
     "chalk": "^2.0.1",
+    "diff-sequences": "^24.3.0",
     "jest-diff": "^24.8.0",
     "jest-get-type": "^24.8.0",
     "pretty-format": "^24.8.0"
diff --git a/packages/jest-matcher-utils/src/cleanupSemantic.ts b/packages/jest-matcher-utils/src/cleanupSemantic.ts
new file mode 100644
index 000000000000..0df74e0bc895
--- /dev/null
+++ b/packages/jest-matcher-utils/src/cleanupSemantic.ts
@@ -0,0 +1,556 @@
+/**
+ * Diff Match and Patch
+ * Copyright 2018 The diff-match-patch Authors.
+ * https://github.com/google/diff-match-patch
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *   http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+/**
+ * @fileoverview Computes the difference between two texts to create a patch.
+ * Applies the patch onto another text, allowing for errors.
+ * @author fraser@google.com (Neil Fraser)
+ */
+
+/**
+ * CHANGES by pedrottimark to diff_match_patch_uncompressed.ts file:
+ *
+ * 1. Delete anything not needed to use diff_cleanupSemantic method
+ * 2. Convert from prototype properties to var declarations
+ * 3. Convert Diff to class from constructor and prototype
+ * 4. Add type annotations for arguments and return values
+ * 5. Add exports
+ */
+
+/**
+ * The data structure representing a diff is an array of tuples:
+ * [[DIFF_DELETE, 'Hello'], [DIFF_INSERT, 'Goodbye'], [DIFF_EQUAL, ' world.']]
+ * which means: delete 'Hello', add 'Goodbye' and keep ' world.'
+ */
+var DIFF_DELETE = -1;
+var DIFF_INSERT = 1;
+var DIFF_EQUAL = 0;
+
+/**
+ * Class representing one diff tuple.
+ * Attempts to look like a two-element array (which is what this used to be).
+ * @param {number} op Operation, one of: DIFF_DELETE, DIFF_INSERT, DIFF_EQUAL.
+ * @param {string} text Text to be deleted, inserted, or retained.
+ * @constructor
+ */
+class Diff {
+  0: number;
+  1: string;
+
+  constructor(op: number, text: string) {
+    this[0] = op;
+    this[1] = text;
+  }
+}
+
+
+/**
+ * Determine the common prefix of two strings.
+ * @param {string} text1 First string.
+ * @param {string} text2 Second string.
+ * @return {number} The number of characters common to the start of each
+ *     string.
+ */
+var diff_commonPrefix = function(text1: string, text2: string): number {
+  // Quick check for common null cases.
+  if (!text1 || !text2 || text1.charAt(0) != text2.charAt(0)) {
+    return 0;
+  }
+  // Binary search.
+  // Performance analysis: https://neil.fraser.name/news/2007/10/09/
+  var pointermin = 0;
+  var pointermax = Math.min(text1.length, text2.length);
+  var pointermid = pointermax;
+  var pointerstart = 0;
+  while (pointermin < pointermid) {
+    if (text1.substring(pointerstart, pointermid) ==
+        text2.substring(pointerstart, pointermid)) {
+      pointermin = pointermid;
+      pointerstart = pointermin;
+    } else {
+      pointermax = pointermid;
+    }
+    pointermid = Math.floor((pointermax - pointermin) / 2 + pointermin);
+  }
+  return pointermid;
+};
+
+
+/**
+ * Determine the common suffix of two strings.
+ * @param {string} text1 First string.
+ * @param {string} text2 Second string.
+ * @return {number} The number of characters common to the end of each string.
+ */
+var diff_commonSuffix = function(text1: string, text2: string): number {
+  // Quick check for common null cases.
+  if (!text1 || !text2 ||
+      text1.charAt(text1.length - 1) != text2.charAt(text2.length - 1)) {
+    return 0;
+  }
+  // Binary search.
+  // Performance analysis: https://neil.fraser.name/news/2007/10/09/
+  var pointermin = 0;
+  var pointermax = Math.min(text1.length, text2.length);
+  var pointermid = pointermax;
+  var pointerend = 0;
+  while (pointermin < pointermid) {
+    if (text1.substring(text1.length - pointermid, text1.length - pointerend) ==
+        text2.substring(text2.length - pointermid, text2.length - pointerend)) {
+      pointermin = pointermid;
+      pointerend = pointermin;
+    } else {
+      pointermax = pointermid;
+    }
+    pointermid = Math.floor((pointermax - pointermin) / 2 + pointermin);
+  }
+  return pointermid;
+};
+
+
+/**
+ * Determine if the suffix of one string is the prefix of another.
+ * @param {string} text1 First string.
+ * @param {string} text2 Second string.
+ * @return {number} The number of characters common to the end of the first
+ *     string and the start of the second string.
+ * @private
+ */
+var diff_commonOverlap_ = function(text1: string, text2: string): number {
+  // Cache the text lengths to prevent multiple calls.
+  var text1_length = text1.length;
+  var text2_length = text2.length;
+  // Eliminate the null case.
+  if (text1_length == 0 || text2_length == 0) {
+    return 0;
+  }
+  // Truncate the longer string.
+  if (text1_length > text2_length) {
+    text1 = text1.substring(text1_length - text2_length);
+  } else if (text1_length < text2_length) {
+    text2 = text2.substring(0, text1_length);
+  }
+  var text_length = Math.min(text1_length, text2_length);
+  // Quick check for the worst case.
+  if (text1 == text2) {
+    return text_length;
+  }
+
+  // Start by looking for a single character match
+  // and increase length until no match is found.
+  // Performance analysis: https://neil.fraser.name/news/2010/11/04/
+  var best = 0;
+  var length = 1;
+  while (true) {
+    var pattern = text1.substring(text_length - length);
+    var found = text2.indexOf(pattern);
+    if (found == -1) {
+      return best;
+    }
+    length += found;
+    if (found == 0 || text1.substring(text_length - length) ==
+        text2.substring(0, length)) {
+      best = length;
+      length++;
+    }
+  }
+};
+
+
+/**
+ * Reduce the number of edits by eliminating semantically trivial equalities.
+ * @param {!Array.<!diff_match_patch.Diff>} diffs Array of diff tuples.
+ */
+ var diff_cleanupSemantic = function(diffs: Array<Diff>) {
+  var changes = false;
+  var equalities = [];  // Stack of indices where equalities are found.
+  var equalitiesLength = 0;  // Keeping our own length var is faster in JS.
+  /** @type {?string} */
+  var lastEquality = null;
+  // Always equal to diffs[equalities[equalitiesLength - 1]][1]
+  var pointer = 0;  // Index of current position.
+  // Number of characters that changed prior to the equality.
+  var length_insertions1 = 0;
+  var length_deletions1 = 0;
+  // Number of characters that changed after the equality.
+  var length_insertions2 = 0;
+  var length_deletions2 = 0;
+  while (pointer < diffs.length) {
+    if (diffs[pointer][0] == DIFF_EQUAL) {  // Equality found.
+      equalities[equalitiesLength++] = pointer;
+      length_insertions1 = length_insertions2;
+      length_deletions1 = length_deletions2;
+      length_insertions2 = 0;
+      length_deletions2 = 0;
+      lastEquality = diffs[pointer][1];
+    } else {  // An insertion or deletion.
+      if (diffs[pointer][0] == DIFF_INSERT) {
+        length_insertions2 += diffs[pointer][1].length;
+      } else {
+        length_deletions2 += diffs[pointer][1].length;
+      }
+      // Eliminate an equality that is smaller or equal to the edits on both
+      // sides of it.
+      if (lastEquality && (lastEquality.length <=
+          Math.max(length_insertions1, length_deletions1)) &&
+          (lastEquality.length <= Math.max(length_insertions2,
+                                           length_deletions2))) {
+        // Duplicate record.
+        diffs.splice(equalities[equalitiesLength - 1], 0,
+                     new Diff(DIFF_DELETE, lastEquality));
+        // Change second copy to insert.
+        diffs[equalities[equalitiesLength - 1] + 1][0] = DIFF_INSERT;
+        // Throw away the equality we just deleted.
+        equalitiesLength--;
+        // Throw away the previous equality (it needs to be reevaluated).
+        equalitiesLength--;
+        pointer = equalitiesLength > 0 ? equalities[equalitiesLength - 1] : -1;
+        length_insertions1 = 0;  // Reset the counters.
+        length_deletions1 = 0;
+        length_insertions2 = 0;
+        length_deletions2 = 0;
+        lastEquality = null;
+        changes = true;
+      }
+    }
+    pointer++;
+  }
+
+  // Normalize the diff.
+  if (changes) {
+    diff_cleanupMerge(diffs);
+  }
+  diff_cleanupSemanticLossless(diffs);
+
+  // Find any overlaps between deletions and insertions.
+  // e.g: <del>abcxxx</del><ins>xxxdef</ins>
+  //   -> <del>abc</del>xxx<ins>def</ins>
+  // e.g: <del>xxxabc</del><ins>defxxx</ins>
+  //   -> <ins>def</ins>xxx<del>abc</del>
+  // Only extract an overlap if it is as big as the edit ahead or behind it.
+  pointer = 1;
+  while (pointer < diffs.length) {
+    if (diffs[pointer - 1][0] == DIFF_DELETE &&
+        diffs[pointer][0] == DIFF_INSERT) {
+      var deletion = diffs[pointer - 1][1];
+      var insertion = diffs[pointer][1];
+      var overlap_length1 = diff_commonOverlap_(deletion, insertion);
+      var overlap_length2 = diff_commonOverlap_(insertion, deletion);
+      if (overlap_length1 >= overlap_length2) {
+        if (overlap_length1 >= deletion.length / 2 ||
+            overlap_length1 >= insertion.length / 2) {
+          // Overlap found.  Insert an equality and trim the surrounding edits.
+          diffs.splice(pointer, 0, new Diff(DIFF_EQUAL,
+              insertion.substring(0, overlap_length1)));
+          diffs[pointer - 1][1] =
+              deletion.substring(0, deletion.length - overlap_length1);
+          diffs[pointer + 1][1] = insertion.substring(overlap_length1);
+          pointer++;
+        }
+      } else {
+        if (overlap_length2 >= deletion.length / 2 ||
+            overlap_length2 >= insertion.length / 2) {
+          // Reverse overlap found.
+          // Insert an equality and swap and trim the surrounding edits.
+          diffs.splice(pointer, 0, new Diff(DIFF_EQUAL,
+              deletion.substring(0, overlap_length2)));
+          diffs[pointer - 1][0] = DIFF_INSERT;
+          diffs[pointer - 1][1] =
+              insertion.substring(0, insertion.length - overlap_length2);
+          diffs[pointer + 1][0] = DIFF_DELETE;
+          diffs[pointer + 1][1] =
+              deletion.substring(overlap_length2);
+          pointer++;
+        }
+      }
+      pointer++;
+    }
+    pointer++;
+  }
+};
+
+
+/**
+ * Look for single edits surrounded on both sides by equalities
+ * which can be shifted sideways to align the edit to a word boundary.
+ * e.g: The c<ins>at c</ins>ame. -> The <ins>cat </ins>came.
+ * @param {!Array.<!diff_match_patch.Diff>} diffs Array of diff tuples.
+ */
+var diff_cleanupSemanticLossless = function(diffs: Array<Diff>) {
+  /**
+   * Given two strings, compute a score representing whether the internal
+   * boundary falls on logical boundaries.
+   * Scores range from 6 (best) to 0 (worst).
+   * Closure, but does not reference any external variables.
+   * @param {string} one First string.
+   * @param {string} two Second string.
+   * @return {number} The score.
+   * @private
+   */
+  function diff_cleanupSemanticScore_(one: string, two: string): number {
+    if (!one || !two) {
+      // Edges are the best.
+      return 6;
+    }
+
+    // Each port of this function behaves slightly differently due to
+    // subtle differences in each language's definition of things like
+    // 'whitespace'.  Since this function's purpose is largely cosmetic,
+    // the choice has been made to use each language's native features
+    // rather than force total conformity.
+    var char1 = one.charAt(one.length - 1);
+    var char2 = two.charAt(0);
+    var nonAlphaNumeric1 = char1.match(nonAlphaNumericRegex_);
+    var nonAlphaNumeric2 = char2.match(nonAlphaNumericRegex_);
+    var whitespace1 = nonAlphaNumeric1 &&
+        char1.match(whitespaceRegex_);
+    var whitespace2 = nonAlphaNumeric2 &&
+        char2.match(whitespaceRegex_);
+    var lineBreak1 = whitespace1 &&
+        char1.match(linebreakRegex_);
+    var lineBreak2 = whitespace2 &&
+        char2.match(linebreakRegex_);
+    var blankLine1 = lineBreak1 &&
+        one.match(blanklineEndRegex_);
+    var blankLine2 = lineBreak2 &&
+        two.match(blanklineStartRegex_);
+
+    if (blankLine1 || blankLine2) {
+      // Five points for blank lines.
+      return 5;
+    } else if (lineBreak1 || lineBreak2) {
+      // Four points for line breaks.
+      return 4;
+    } else if (nonAlphaNumeric1 && !whitespace1 && whitespace2) {
+      // Three points for end of sentences.
+      return 3;
+    } else if (whitespace1 || whitespace2) {
+      // Two points for whitespace.
+      return 2;
+    } else if (nonAlphaNumeric1 || nonAlphaNumeric2) {
+      // One point for non-alphanumeric.
+      return 1;
+    }
+    return 0;
+  }
+
+  var pointer = 1;
+  // Intentionally ignore the first and last element (don't need checking).
+  while (pointer < diffs.length - 1) {
+    if (diffs[pointer - 1][0] == DIFF_EQUAL &&
+        diffs[pointer + 1][0] == DIFF_EQUAL) {
+      // This is a single edit surrounded by equalities.
+      var equality1 = diffs[pointer - 1][1];
+      var edit = diffs[pointer][1];
+      var equality2 = diffs[pointer + 1][1];
+
+      // First, shift the edit as far left as possible.
+      var commonOffset = diff_commonSuffix(equality1, edit);
+      if (commonOffset) {
+        var commonString = edit.substring(edit.length - commonOffset);
+        equality1 = equality1.substring(0, equality1.length - commonOffset);
+        edit = commonString + edit.substring(0, edit.length - commonOffset);
+        equality2 = commonString + equality2;
+      }
+
+      // Second, step character by character right, looking for the best fit.
+      var bestEquality1 = equality1;
+      var bestEdit = edit;
+      var bestEquality2 = equality2;
+      var bestScore = diff_cleanupSemanticScore_(equality1, edit) +
+          diff_cleanupSemanticScore_(edit, equality2);
+      while (edit.charAt(0) === equality2.charAt(0)) {
+        equality1 += edit.charAt(0);
+        edit = edit.substring(1) + equality2.charAt(0);
+        equality2 = equality2.substring(1);
+        var score = diff_cleanupSemanticScore_(equality1, edit) +
+            diff_cleanupSemanticScore_(edit, equality2);
+        // The >= encourages trailing rather than leading whitespace on edits.
+        if (score >= bestScore) {
+          bestScore = score;
+          bestEquality1 = equality1;
+          bestEdit = edit;
+          bestEquality2 = equality2;
+        }
+      }
+
+      if (diffs[pointer - 1][1] != bestEquality1) {
+        // We have an improvement, save it back to the diff.
+        if (bestEquality1) {
+          diffs[pointer - 1][1] = bestEquality1;
+        } else {
+          diffs.splice(pointer - 1, 1);
+          pointer--;
+        }
+        diffs[pointer][1] = bestEdit;
+        if (bestEquality2) {
+          diffs[pointer + 1][1] = bestEquality2;
+        } else {
+          diffs.splice(pointer + 1, 1);
+          pointer--;
+        }
+      }
+    }
+    pointer++;
+  }
+};
+
+
+// Define some regex patterns for matching boundaries.
+var nonAlphaNumericRegex_ = /[^a-zA-Z0-9]/;
+var whitespaceRegex_ = /\s/;
+var linebreakRegex_ = /[\r\n]/;
+var blanklineEndRegex_ = /\n\r?\n$/;
+var blanklineStartRegex_ = /^\r?\n\r?\n/;
+
+
+/**
+ * Reorder and merge like edit sections.  Merge equalities.
+ * Any edit section can move as long as it doesn't cross an equality.
+ * @param {!Array.<!diff_match_patch.Diff>} diffs Array of diff tuples.
+ */
+var diff_cleanupMerge = function(diffs: Array<Diff>) {
+  // Add a dummy entry at the end.
+  diffs.push(new Diff(DIFF_EQUAL, ''));
+  var pointer = 0;
+  var count_delete = 0;
+  var count_insert = 0;
+  var text_delete = '';
+  var text_insert = '';
+  var commonlength;
+  while (pointer < diffs.length) {
+    switch (diffs[pointer][0]) {
+      case DIFF_INSERT:
+        count_insert++;
+        text_insert += diffs[pointer][1];
+        pointer++;
+        break;
+      case DIFF_DELETE:
+        count_delete++;
+        text_delete += diffs[pointer][1];
+        pointer++;
+        break;
+      case DIFF_EQUAL:
+        // Upon reaching an equality, check for prior redundancies.
+        if (count_delete + count_insert > 1) {
+          if (count_delete !== 0 && count_insert !== 0) {
+            // Factor out any common prefixies.
+            commonlength = diff_commonPrefix(text_insert, text_delete);
+            if (commonlength !== 0) {
+              if ((pointer - count_delete - count_insert) > 0 &&
+                  diffs[pointer - count_delete - count_insert - 1][0] ==
+                  DIFF_EQUAL) {
+                diffs[pointer - count_delete - count_insert - 1][1] +=
+                    text_insert.substring(0, commonlength);
+              } else {
+                diffs.splice(0, 0, new Diff(DIFF_EQUAL,
+                    text_insert.substring(0, commonlength)));
+                pointer++;
+              }
+              text_insert = text_insert.substring(commonlength);
+              text_delete = text_delete.substring(commonlength);
+            }
+            // Factor out any common suffixies.
+            commonlength = diff_commonSuffix(text_insert, text_delete);
+            if (commonlength !== 0) {
+              diffs[pointer][1] = text_insert.substring(text_insert.length -
+                  commonlength) + diffs[pointer][1];
+              text_insert = text_insert.substring(0, text_insert.length -
+                  commonlength);
+              text_delete = text_delete.substring(0, text_delete.length -
+                  commonlength);
+            }
+          }
+          // Delete the offending records and add the merged ones.
+          pointer -= count_delete + count_insert;
+          diffs.splice(pointer, count_delete + count_insert);
+          if (text_delete.length) {
+            diffs.splice(pointer, 0,
+                new Diff(DIFF_DELETE, text_delete));
+            pointer++;
+          }
+          if (text_insert.length) {
+            diffs.splice(pointer, 0,
+                new Diff(DIFF_INSERT, text_insert));
+            pointer++;
+          }
+          pointer++;
+        } else if (pointer !== 0 && diffs[pointer - 1][0] == DIFF_EQUAL) {
+          // Merge this equality with the previous one.
+          diffs[pointer - 1][1] += diffs[pointer][1];
+          diffs.splice(pointer, 1);
+        } else {
+          pointer++;
+        }
+        count_insert = 0;
+        count_delete = 0;
+        text_delete = '';
+        text_insert = '';
+        break;
+    }
+  }
+  if (diffs[diffs.length - 1][1] === '') {
+    diffs.pop();  // Remove the dummy entry at the end.
+  }
+
+  // Second pass: look for single edits surrounded on both sides by equalities
+  // which can be shifted sideways to eliminate an equality.
+  // e.g: A<ins>BA</ins>C -> <ins>AB</ins>AC
+  var changes = false;
+  pointer = 1;
+  // Intentionally ignore the first and last element (don't need checking).
+  while (pointer < diffs.length - 1) {
+    if (diffs[pointer - 1][0] == DIFF_EQUAL &&
+        diffs[pointer + 1][0] == DIFF_EQUAL) {
+      // This is a single edit surrounded by equalities.
+      if (diffs[pointer][1].substring(diffs[pointer][1].length -
+          diffs[pointer - 1][1].length) == diffs[pointer - 1][1]) {
+        // Shift the edit over the previous equality.
+        diffs[pointer][1] = diffs[pointer - 1][1] +
+            diffs[pointer][1].substring(0, diffs[pointer][1].length -
+                                        diffs[pointer - 1][1].length);
+        diffs[pointer + 1][1] = diffs[pointer - 1][1] + diffs[pointer + 1][1];
+        diffs.splice(pointer - 1, 1);
+        changes = true;
+      } else if (diffs[pointer][1].substring(0, diffs[pointer + 1][1].length) ==
+          diffs[pointer + 1][1]) {
+        // Shift the edit over the next equality.
+        diffs[pointer - 1][1] += diffs[pointer + 1][1];
+        diffs[pointer][1] =
+            diffs[pointer][1].substring(diffs[pointer + 1][1].length) +
+            diffs[pointer + 1][1];
+        diffs.splice(pointer + 1, 1);
+        changes = true;
+      }
+    }
+    pointer++;
+  }
+  // If shifts were made, the diff needs reordering and another shift sweep.
+  if (changes) {
+    diff_cleanupMerge(diffs);
+  }
+};
+
+
+export {
+  Diff,
+  DIFF_EQUAL,
+  DIFF_DELETE,
+  DIFF_INSERT,
+  diff_cleanupSemantic as cleanupSemantic,
+};
diff --git a/packages/jest-matcher-utils/src/diffStrings.ts b/packages/jest-matcher-utils/src/diffStrings.ts
new file mode 100644
index 000000000000..6df165de718a
--- /dev/null
+++ b/packages/jest-matcher-utils/src/diffStrings.ts
@@ -0,0 +1,58 @@
+/**
+ * Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
+ *
+ * This source code is licensed under the MIT license found in the
+ * LICENSE file in the root directory of this source tree.
+ */
+
+import diffSequences from 'diff-sequences';
+
+import {
+  cleanupSemantic,
+  Diff,
+  DIFF_EQUAL,
+  DIFF_DELETE,
+  DIFF_INSERT,
+} from './cleanupSemantic';
+
+const diffStrings = (a: string, b: string): Array<Diff> | null => {
+  const isCommon = (aIndex: number, bIndex: number) => a[aIndex] === b[bIndex];
+
+  let aIndex = 0;
+  let bIndex = 0;
+  const diffs: Array<Diff> = [];
+
+  const foundSubsequence = (
+    nCommon: number,
+    aCommon: number,
+    bCommon: number,
+  ) => {
+    if (aIndex !== aCommon) {
+      diffs.push(new Diff(DIFF_DELETE, a.slice(aIndex, aCommon)));
+    }
+    if (bIndex !== bCommon) {
+      diffs.push(new Diff(DIFF_INSERT, b.slice(bIndex, bCommon)));
+    }
+
+    aIndex = aCommon + nCommon; // number of characters compared in a
+    bIndex = bCommon + nCommon; // number of characters compared in b
+    diffs.push(new Diff(DIFF_EQUAL, b.slice(bCommon, bIndex)));
+  };
+
+  diffSequences(a.length, b.length, isCommon, foundSubsequence);
+
+  // After the last common subsequence, push remaining change items.
+  if (aIndex !== a.length) {
+    diffs.push(new Diff(DIFF_DELETE, a.slice(aIndex)));
+  }
+  if (bIndex !== b.length) {
+    diffs.push(new Diff(DIFF_INSERT, b.slice(bIndex)));
+  }
+
+  cleanupSemantic(diffs);
+
+  // Assume it has a change string, but does it have a common string?
+  return diffs.some(diff => diff[0] === DIFF_EQUAL) ? diffs : null;
+};
+
+export default diffStrings;
diff --git a/packages/jest-matcher-utils/src/index.ts b/packages/jest-matcher-utils/src/index.ts
index 9f9de13390ce..ee0b59edc999 100644
--- a/packages/jest-matcher-utils/src/index.ts
+++ b/packages/jest-matcher-utils/src/index.ts
@@ -7,8 +7,16 @@
 
 import chalk from 'chalk';
 import jestDiff, {DiffOptions} from 'jest-diff';
-import getType from 'jest-get-type';
+import getType, {isPrimitive} from 'jest-get-type';
 import prettyFormat from 'pretty-format';
+
+import diffStrings from './diffStrings';
+import {
+  MULTILINE_REGEXP,
+  getExpectedString,
+  getReceivedString,
+} from './printDiff';
+
 const {
   AsymmetricMatcher,
   DOMCollection,
@@ -198,6 +206,112 @@ export const ensureExpectedIsNonNegativeInteger = (
   }
 };
 
+const isDiffable = (expected: unknown, received: unknown): boolean => {
+  const expectedType = getType(expected);
+  const receivedType = getType(received);
+
+  if (expectedType !== receivedType) {
+    return false;
+  }
+
+  if (isPrimitive(expected)) {
+    // Print diff only if both strings have more than one line.
+    return (
+      typeof expected === 'string' &&
+      typeof received === 'string' &&
+      MULTILINE_REGEXP.test(expected) &&
+      MULTILINE_REGEXP.test(received)
+    );
+  }
+
+  if (
+    expectedType === 'date' ||
+    expectedType === 'function' ||
+    expectedType === 'regexp'
+  ) {
+    return false;
+  }
+
+  if (expected instanceof Error && received instanceof Error) {
+    return false;
+  }
+
+  if (
+    expectedType === 'object' &&
+    typeof (expected as any).asymmetricMatch === 'function'
+  ) {
+    return false;
+  }
+
+  if (
+    receivedType === 'object' &&
+    typeof (received as any).asymmetricMatch === 'function'
+  ) {
+    return false;
+  }
+
+  return true;
+};
+
+export const printDiffOrStringify = (
+  expected: unknown,
+  received: unknown,
+  expectedLabel: string,
+  receivedLabel: string,
+  expand: boolean, // CLI options: true if `--expand` or false if `--no-expand`
+): string => {
+  const printLabel = getLabelPrinter(expectedLabel, receivedLabel);
+
+  if (
+    typeof expected === 'string' &&
+    typeof received === 'string' &&
+    expected.length !== 0 &&
+    received.length !== 0 &&
+    !MULTILINE_REGEXP.test(expected) &&
+    !MULTILINE_REGEXP.test(received)
+  ) {
+    const diffs = diffStrings(expected, received);
+
+    if (Array.isArray(diffs)) {
+      const expectedLine =
+        printLabel(expectedLabel) + printExpected(getExpectedString(diffs));
+      const receivedLine =
+        printLabel(receivedLabel) + printReceived(getReceivedString(diffs));
+
+      return expectedLine + '\n' + receivedLine;
+    }
+  } else if (isDiffable(expected, received)) {
+    // Cannot use same serialization as shortcut to avoid diff,
+    // because stringify (that is, pretty-format with min option)
+    // omits constructor name for array or object, too bad so sad :(
+    const difference = jestDiff(expected, received, {
+      aAnnotation: expectedLabel,
+      bAnnotation: receivedLabel,
+      expand,
+    });
+
+    if (
+      typeof difference === 'string' &&
+      difference.includes('- ' + expectedLabel) &&
+      difference.includes('+ ' + receivedLabel)
+    ) {
+      return difference;
+    }
+  }
+
+  // Cannot reuse value of stringify(received) in report string,
+  // because printReceived does inverse highlight space at end of line,
+  // but RECEIVED_COLOR does not (it refers to a plain chalk method).
+  return (
+    `${printLabel(expectedLabel)}${printExpected(expected)}\n` +
+    `${printLabel(receivedLabel)}${
+      stringify(expected) === stringify(received)
+        ? 'serializes to the same string'
+        : printReceived(received)
+    }`
+  );
+};
+
 // Sometimes, e.g. when comparing two numbers, the output from jest-diff
 // does not contain more information than the `Expected:` / `Received:` already gives.
 // In those cases, we do not print a diff to make the output shorter and not redundant.
diff --git a/packages/jest-matcher-utils/src/printDiff.ts b/packages/jest-matcher-utils/src/printDiff.ts
new file mode 100644
index 000000000000..6866d8513f80
--- /dev/null
+++ b/packages/jest-matcher-utils/src/printDiff.ts
@@ -0,0 +1,36 @@
+/**
+ * Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
+ *
+ * This source code is licensed under the MIT license found in the
+ * LICENSE file in the root directory of this source tree.
+ */
+
+import {Diff, DIFF_EQUAL, DIFF_DELETE, DIFF_INSERT} from './cleanupSemantic';
+import {INVERTED_COLOR} from './index';
+
+export const getDiffString = (diffs: Array<Diff>, op: number): string => {
+  const hasEqual = diffs.some(
+    diff => diff[0] === DIFF_EQUAL && diff[1].length !== 0,
+  );
+
+  return diffs.reduce(
+    (reduced: string, diff: Diff): string =>
+      reduced +
+      (diff[0] === DIFF_EQUAL
+        ? diff[1]
+        : diff[0] !== op
+        ? ''
+        : hasEqual
+        ? INVERTED_COLOR(diff[1])
+        : diff[1]),
+    '',
+  );
+};
+
+export const getExpectedString = (diffs: Array<Diff>): string =>
+  getDiffString(diffs, DIFF_DELETE);
+
+export const getReceivedString = (diffs: Array<Diff>): string =>
+  getDiffString(diffs, DIFF_INSERT);
+
+export const MULTILINE_REGEXP = /\n/;
diff --git a/packages/jest-matcher-utils/tsconfig.json b/packages/jest-matcher-utils/tsconfig.json
index 2248b5ff7fa1..8210e0546bad 100644
--- a/packages/jest-matcher-utils/tsconfig.json
+++ b/packages/jest-matcher-utils/tsconfig.json
@@ -5,6 +5,7 @@
     "outDir": "build"
   },
   "references": [
+    {"path": "../diff-sequences"},
     {"path": "../jest-diff"},
     {"path": "../jest-get-type"},
     {"path": "../pretty-format"}
diff --git a/scripts/checkCopyrightHeaders.js b/scripts/checkCopyrightHeaders.js
index e7e1cc9e7dd7..17587bb26337 100755
--- a/scripts/checkCopyrightHeaders.js
+++ b/scripts/checkCopyrightHeaders.js
@@ -101,6 +101,7 @@ const CUSTOM_IGNORED_PATTERNS = [
   '^flow-typed/.*',
   '^packages/expect/src/jasmineUtils\\.ts$',
   '^packages/jest-config/src/vendor/jsonlint\\.js$',
+  '^packages/jest-matcher-utils/src/cleanupSemantic\\.ts$',
 ].map(createRegExp);
 
 const IGNORED_PATTERNS = [