Defining sample size in attribute control charts (ACC) is a problem, though it is one of the important techniques in process control. Solution of this problem is investigated in this paper. For this purpose, a model is developed in multistage processes and it is solved by genetic algorithms (GAs). The aim of the model is to determine the sample size for ACC. The model has two main goals: first is to determine the best acceptance probability and second is to provide the minimum cost in every stage. Formulations of this model are calculated based on acceptance sampling approach. When GA solves the model, two main parameters are determined for every stage. These are: sample size, n, and acceptance number, c. This approach is applied in an engine piston manufacturing firm and sample size, n, which is determined in model by GA, is suggested for ACC. (c) 2006 Elsevier B.V. All rights reserved.