This article presents a numerical alternative for treating the propagation of uncertainty in a function that depends on random experimental variables. With the help of microcomputers, this technique can be used by students at all levels in a college physics curriculum and is particularly useful when there are a large number of such calculations to do or the function involved is complicated due to the number of variables present or the explicit functional form. The numerical procedure can even be used with a function that cannot be expressed analytically, but is given instead in the form of a table or graph. Finally, the technique does not require knowledge of partial differentiation or multivariable Taylor series expansions in contrast to the traditional treatment of the propagation of uncertainty.