diff --git a/pkg/local_object_storage/metabase/metadata.go b/pkg/local_object_storage/metabase/metadata.go
index 960ebf9c04..b8612cf899 100644
--- a/pkg/local_object_storage/metabase/metadata.go
+++ b/pkg/local_object_storage/metabase/metadata.go
@@ -312,7 +312,6 @@ func (db *DB) searchTx(tx *bbolt.Tx, cnr cid.ID, fs object.SearchFilters, fInt m
 	idIter := len(attrs) == 0 || notPresentPrimMatcher
 	primAttr := fs[0].Header() // attribute emptiness already prevented
 	var primSeekKey, primSeekPrefix []byte
-	var prevResOID, prevResPrimVal []byte
 	if idIter {
 		if cursor != nil {
 			primSeekKey = cursor.Key
@@ -329,15 +328,6 @@ func (db *DB) searchTx(tx *bbolt.Tx, cnr cid.ID, fs object.SearchFilters, fInt m
 			primSeekKey[0] = metaPrefixAttrIDPlain
 		}
 		primSeekPrefix = primSeekKey[:cursor.ValIDOff]
-		valID := cursor.Key[cursor.ValIDOff:]
-		if intPrimMatcher {
-			prevResPrimVal, prevResOID = valID[:len(valID)-oid.Size], valID[len(valID)-oid.Size:]
-		} else {
-			var err error
-			if prevResPrimVal, prevResOID, err = splitValOID(valID); err != nil {
-				return nil, nil, fmt.Errorf("invalid VAL_OID: %w", err)
-			}
-		}
 	} else {
 		if intPrimMatcher {
 			// we seek 0x01_ATTR_DELIM_VAL either w/ or w/o VAL. We ignore VAL when we need
@@ -379,8 +369,7 @@ func (db *DB) searchTx(tx *bbolt.Tx, cnr cid.ID, fs object.SearchFilters, fInt m
 	}
 
 	res := make([]client.SearchResultItem, count)
-	collectedPrimVals := make([][]byte, count)
-	collectedPrimKeys := make([][]byte, count) // TODO: can be done w/o slice
+	var lastMatchedPrimKey []byte
 	var n uint16
 	var more bool
 	var id, dbVal, primDBVal []byte
@@ -488,26 +477,14 @@ nextPrimKey:
 		if objectStatus(tx, oid.NewAddress(cnr, oid.ID(id)), curEpoch) > 0 { // GC-ed
 			continue nextPrimKey
 		}
-		// object matches, collect attributes
+		// object matches
+		if n == count {
+			more = true
+			break
+		}
+		// collect attributes
 		collected := make([]string, len(attrs))
-		var insertI uint16
 		if len(attrs) > 0 {
-			if cursor != nil { // can be < than previous response chunk
-				if c := bytes.Compare(primDBVal, prevResPrimVal); c < 0 || c == 0 && bytes.Compare(id, prevResOID) <= 0 {
-					continue nextPrimKey
-				}
-				// note that if both values are integers, they are already sorted. Otherwise, the order is undefined.
-				// We could treat non-int values as < then the int ones, but the code would have grown huge
-			}
-			for i := range n {
-				if c := bytes.Compare(primDBVal, collectedPrimVals[i]); c < 0 || c == 0 && bytes.Compare(id, res[i].ID[:]) < 0 {
-					break
-				}
-				if insertI++; insertI == count {
-					more = true
-					continue nextPrimKey
-				}
-			}
 			if intPrimMatcher {
 				var err error
 				if collected[0], err = restoreIntAttribute(primDBVal); err != nil {
@@ -517,19 +494,6 @@ nextPrimKey:
 				collected[0] = string(primDBVal)
 			}
 		} else {
-			if cursor != nil { // can be < than previous response chunk
-				if bytes.Compare(id, prevResOID) <= 0 {
-					continue nextPrimKey
-				}
-			}
-			for i := insertI; i < n; i++ {
-				if bytes.Compare(id, res[i].ID[:]) >= 0 {
-					if insertI++; insertI == count {
-						more = true
-						continue nextPrimKey
-					}
-				}
-			}
 		}
 		for i := 1; i < len(attrs); i++ {
 			val, err := attrSkr.get(id, attrs[i])
@@ -540,26 +504,16 @@ nextPrimKey:
 				return nil, nil, err
 			}
 		}
-		if n == count {
-			more = true
-			if len(attrs) > 0 {
-				break
-			} // else, for empty attrs, "later" objects may have "less" ID, so we should continue
-		}
-		copy(res[insertI+1:], res[insertI:])
-		res[insertI].ID = oid.ID(id)
-		res[insertI].Attributes = collected
-		copy(collectedPrimVals[insertI+1:], collectedPrimVals[insertI:])
-		collectedPrimVals[insertI] = primDBVal
-		copy(collectedPrimKeys[insertI+1:], collectedPrimKeys[insertI:])
-		collectedPrimKeys[insertI] = primKey
-		if n < count {
-			n++
-		}
+		res[n].ID = oid.ID(id)
+		res[n].Attributes = collected
+		lastMatchedPrimKey = primKey
+		n++
+		// note that even when n == count meaning there will be no more attachments, we
+		// still need to determine whether there more matching elements (see condition above)
 	}
 	var newCursor []byte
 	if more {
-		newCursor = slices.Clone(collectedPrimKeys[n-1][1:])
+		newCursor = slices.Clone(lastMatchedPrimKey[1:])
 	}
 	return res[:n], newCursor, nil
 }