Cylindrical feature has become one of the most important features in mechanical designs. It is essential to establish a proper model for evaluating cylindricity error to ensure achieving accurate assessment based on the new generation geometrical product specification (GPS). In this paper we developed a method for uncertainty evaluation of cylindricity errors. This method is based on the particle swarm optimization (PSO) algorithm and the uncertainty theory in new generation GPS to address the issues involved in uncertainty in new GPS system. The association operation in GPS was used as an evaluation method for cylindricity error. One of the advantages of the proposed approach is that the PSO algorithm has faster convergence speed and calculating precision, the evaluation results are more accurate and accord with requirements of new generation GPS standard. We present an example to describe how uncertainty can be used as an economic tool to help commercial transactions where the workpiece is evaluated against the specification. The result indicates that the proposed method can improve the veracity of evaluation.