The Wisconsin card sorting test (WCST) comprehensively examines executive function (EF). The Keio version of the WCST (KWCST) uses fewer cards and presents them in two steps, separated by a short pause during which an instruction is given. Being of short duration, this test is suitable for children with pervasive developmental disorder (PDD) and attention-deficit/hyperactivity disorder (AD/HD), yet few studies have examined the performance of children with such developmental disorders in the second step of a two-step test such as the KWCST. Using the KWCST, this study compares EF in children with PDD (n = 52), or AD/HD (n = 46) to that in typically developing (TD) children (n = 52). Scores for the six indices of this test, including numbers of response cards until the first category achieved (NUCA), total errors (TE), and non-perseverative errors of Nelson (NPEN), were analyzed using ANOVA. Compared to the TD group, scores in the PDD and/or AD/HD groups were significantly lower for all indices except NUCA and NPEN for the first step, and lower for all indices except NUCA for the second step. Moreover, significantly fewer improvements in TE were seen in the PDD group, and significantly fewer improvements in NPEN were seen in the AD/HD group, compared with TD. This study suggests that both PDD and AD/HD make it difficult for children to utilize their experience in the first step and to effectively apply the instruction given before the second step. It also suggests that the two-step nature of the KWCST is clinically important. (C) 2011 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.