A method of representing numbers in which the decimal or binary point is fixed in a given location and the number contains a constant predetermined number of digits. Compare floating-point representation.