The criterion problem in the neoPiagetian cognitive-developmental literature is reviewed. This problem has provoked debate in the context of three empirical questions: concept age norms, concept training effects, and concept invariant sequences. It is argued that only the question of invaniant sequences in same-stage concepts is theoretically crucial. With regard to this question it is shown that both Type I ("false positive") and Type II ("false negative") criterion errors tend to produce data which are spuriously supportive of theoretical predictions. It is concluded that the objective psychometric consequences of criterion errors provide no grounds for the current view that obtaining or not obtaining theory-supportive findings in concept development research turns on arbitrary decisions about response criteria. (Author/CS)