In this article an economic analysis is presented for the problem of selecting the most favorable quality distribution for an industrial process. An important special case is that of selecting the best mean (target) value. The analysis, which takes into account the stochastic nature of the process, is illustrated by considering the problem of choosing the optimal amount of overfill in a filling operation. A detailed solution is given for situations in which the distribution of the quality measurements is approximately normal; for that case a table is provided that simplifies the solution of the problem in practice. Industrial examples are considered. Lognormal and Poisson distributions are also discussed.