Skinner's method provides a means of computing; numerically, upper and lower bounds on a cumulative distribution function resulting from the convolution of probability density functions. The method thus provides approximate numerical results whose accuracy is known precisely. This paper provides an exposition of Skinner's method. It shows how the method Can be applied to the computation of numerical solutions of other problems, as well as the waiting time distribution of theM/G/1queue, for which Skinner presented the method.