This paper presents the theoretical developments of a novel approach for the optimization of design models involving stochastic parameters. Based on the postulation of general probability distribution functions describing process uncertainty and variations, the problem is formulated as a two-stage stochastic programming problem where the objective is to determine the design that maximizes an expected revenue or profit while simultaneously measuring design feasibility. In contrast to previous approaches, design feasibility and economic optimality are simultaneously obtained without requiring an a priori’ discretization of the uncertainty. The decomposition based approach is general to deal with linear and convex nonlinear models involving uncertainty described by arbitrary probability distribution functions. A very attractive feature of this approach is that since most optimization tasks can be performed independently the algorithm has a highly parallel structure that can be further exploited. The steps of the proposed algorithmic procedure are illustrated in detail through four example problems.