Cognitive control is required for flexible responses in changing environments and varies in its effectiveness even among healthy people. We evaluated cognitive control using antisaccades in comparison with basic prosaccades. Documented relationships exist between antisaccade error rate and prosaccade latency: individuals with shorter prosaccade latency show more antisaccade errors. Previous studies also suggest that individual differences in working memory may influence saccade performance. The current study investigated the relationships among prosaccade, antisaccade and working memory (assessed by symmetry span) data collected from over 150 healthy young adults. Hierarchical multiple regression analyses were conducted. The results demonstrated that prosaccade latency was a reliable predictor of antisaccade error rate, and working memory moderated the predictability of the two. These results suggested that working memory may contribute strongly to the individual differences observed with respect to differences in cognitive control among healthy people.