diff --git a/docs/contributing.md b/docs/contributing.md
index 0559b2cf9..dd2dc6019 100644
--- a/docs/contributing.md
+++ b/docs/contributing.md
@@ -58,7 +58,7 @@ Each pull request should include:
  - The implementation of the new metric or score in xarray, ideally with support for pandas and dask.
  - 100% unit test coverage.
  - A tutorial notebook showcasing the use of that metric or score, ideally based on the standard sample data.
- - API documentation (docstrings) using [Napoleon (google)](https://sphinxcontrib-napoleon.readthedocs.io/en/latest/example_google.html) style, making sure to clearly explain the use of the metric.
+ - API documentation (docstrings) using [Napoleon (google)](https://sphinxcontrib-napoleon.readthedocs.io/en/latest/example_google.html) style, making sure to clearly explain the use of the metric. Please see our [docstring template](docstring_template.md).
  - A reference should be added to the API documentation. The preferred referencing style is [APA (7th edition)](https://apastyle.apa.org/style-grammar-guidelines/references/examples).
    - If there is an authoritative reference (e.g. an academic article that defines the metric, or the specific implementation, being used), please cite that reference.
    - When there is an authoritative reference, please cite it even if it cannot be accessed without payment. In these instances, if there is another suitable reference that is open and available without payment, cite that as well.
diff --git a/docs/docstring_template.md b/docs/docstring_template.md
new file mode 100644
index 000000000..92a840d57
--- /dev/null
+++ b/docs/docstring_template.md
@@ -0,0 +1,72 @@
+# Docstring Template
+--------------------
+
+Below is a docstring example. It contains a suggested docstring layout, example text and example MathJax.
+
+The layout below will render properly in Read the Docs. Please adjust the example text and MathJax as required.
+
+Note: backticks render differently in docstrings to elsewhere. Should you wish to include text in backticks, please use double backticks - e.g. ``scores.continuous.mean_error``. (If you only use one set of backticks, the text will instead be rendered as italics in Read the Docs). 
+
+```
+def function_name(*args, *, **kwargs) -> type hint:
+    '''
+    A short one or two line description of what the function does. 
+
+    Additional information (e.g. a paragraph or two) may be included here following the short description.
+
+    MathJax giving the mathematical formula for the metric. Below is some MathJax to use as a 
+    starting point. Note that double-backslash formatting is required to ensure proper rendering in docstrings.
+
+    .. math::
+        \\text{mean error} =\\frac{1}{N}\\sum_{i=1}^{N}(x_i - y_i)
+        \\text{where } x = \\text{the forecast, and } y = \\text{the observation}	
+
+    Where applicable, say what the range of the score is. For example:
+    Range: 0 to infinity, 0 is perfect. 
+
+    Mathjax can also be included inline, like this :math:`\\text{obs_cdf}(x) = 0`
+
+    Args: 
+        fcst: Forecast data. 
+        obs: Observation data.
+        reduce_dims: Dimensions to reduce. Can be "all" to reduce all dimensions.
+        preserve_dims: Dimensions to preserve. Can be "all" to preserve all dimensions.
+        weights: Weights for calculating a weighted mean of individual scores (e.g. by area, by latitude, 
+            by population, custom). Note that these weights are different to threshold weighting which is done 
+            by decision threshold.
+
+    Returns:
+        A semantic description of what is returned. Note - type information is automatically included based on type hinting.
+
+    Raises:
+        A type and description of any special error checking which may result in an exception, e.g.
+        ValueError: if there are values in `fcst` and `obs` which are invalid
+
+    Notes:
+        - This section is optional.
+        - Any additional comments or notes should go here.
+
+    References:
+        - If possible, a citation to the first (original) paper introducing the score/metric.
+        - In addition, if that paper is not open access, please also add an open access reference.
+        - The preferred referencing style for journal articles is [APA (7th edition).](https://apastyle.apa.org/style-grammar-guidelines/references/examples/journal-article-references)
+        - Example reference below:
+        - Sorooshian, S., Duan, Q., & Gupta, V. K. (1993). Calibration of rainfall-runoff models:
+          Application of global optimization to the Sacramento Soil Moisture Accounting Model.
+          Water Resources Research, 29(4), 1185-1194. https://doi.org/10.1029/92WR02617        
+        - Optionally, if there is a website(s) that has particularly good information about the metric, please feel free to list it as well.
+
+    See also:
+        - This section is optional.
+        - If there are other functions which are directly relevant, you can add e.g. :py:func:`scores.continuous.rmse` with a description.
+
+    Examples:
+
+        >>> import numpy as np  # NOTE - the previous line should be empty for rendering to work properly.
+        >>> import xarray as xr
+        >>> from scores.probability import interval_tw_crps_for_ensemble
+        >>> fcst = xr.DataArray(np.random.uniform(-40, 10, size=(10, 10)), dims=['time', 'ensemble'])
+        >>> obs = xr.DataArray(np.random.uniform(-40, 10, size=10), dims=['time'])
+
+    '''
+```
\ No newline at end of file