Response surface methodology is used to model properties of positive photoresist which have been subjected to several operations in a photolithographic process. The impact of these models on process performance and latitude has been investigated and a general approach to process optimization has been proposed: m P[X(1),X(2), . . . 1 = fl F(i) i =0 where P[X(1),X(2), . . .1 represents an overall process optimization function. It measures the overall performance and stability of a process as a func-tion of process variables X(1), X(2), etc. This optimization function is defined as the product of normalized signal-to-noise ratios, F(i), for the set of responses, i, considered. The function F(i) quantifies the ability of a process to achieve the specified response and the sensitivity of the response to perturbations in the process variables. This approach is particularly useful when more than one response must be optimized with a given process. The application of this approach can result in a several-fold increase in process performance and latitude. Examples are presented and discussed.