diff --git a/docs/function-math-truncate.md b/docs/function-math-truncate.md
new file mode 100644
index 000000000..ae2a3c9a3
--- /dev/null
+++ b/docs/function-math-truncate.md
@@ -0,0 +1,79 @@
+# math.truncate
+
+## Summary
+
+A method that returns the integer part of a number by removing any fractional digits. Optionally, along with specifying how many digits should be kept after the decimal point.
+
+## Motivation
+
+``math.floor`` does not give the same effect as ``Math.trunc()`` from JavaScript, for instance. But ``math.floor`` and ``math.ceil`` can be used together, to re-produce the same result that ``Math.trunc()`` would give, but a function like ``math.truncate`` directly as a built-in would be more straightforward.
+
+For instance, negative numbers:
+```lua
+print( math.floor(-0.005) )
+-- Returns: -1
+
+print( math.ceil(-0.005) )
+-- Returns: -0
+```
+
+
+
+```js
+Math.trunc(-0.005)
+// Returns: -0
+```
+
+Truncation specifically just cuts off the digits, while the others try to round.
+
+
+## Design
+
+``math.truncate(num, idp)``
+- ``num``: Number to truncate
+- ``idp``: _(Optional)_, How many digits to keep after the decimal point
+
+
+```lua
+math.truncate(-0.005)
+-- Expected output: -0
+
+math.truncate(0.005)
+-- Expected output: 0
+
+math.truncate(1.23456789, 4)
+-- Expected output: 1.2345
+
+math.truncate(-math.huge)
+-- Expected output: -inf
+```
+
+
+## Drawbacks
+
+- Should ``idp`` be a thing? Since one can just do this _(JavaScript example)_
+
+```js
+function trunc_with_idp(num, idp) {
+    let mul = Math.pow(10, (idp || 0))
+    return Math.trunc(num * mul) / mul
+}
+trunc_with_idp(1.2345, 4)
+```
+
+_or see alternative below_
+
+
+Other questions to ask:
+- Will a built-in offer optimization over the alternatives below?
+
+
+## Alternatives
+
+A Lua function from [here](https://github.com/Facepunch/garrysmod/blob/8e96e31fc9ab9de053edc0513f9d7a6f3e4be1e1/garrysmod/lua/includes/extensions/math.lua#L168-L171)
+```lua
+function math.Truncate( num, idp )
+	local mult = 10 ^ ( idp or 0 )
+	return ( num < 0 and math.ceil or math.floor )( num * mult ) / mult
+end
+```