Previous research suggests that English-speaking children comprehend agent-patient verb passives earlier than experiencer-theme verb passives (Maratsos, Fox, Becker, & Chalkley, 1985). We report three experiments examining whether such effects reflect delayed acquisition of the passive syntax or instead are an artifact of the experimental task, relating to children's poor picture recognition for such verbs. In two syntactic priming experiments, 3- and 4-year-olds produced more agent-patient passives after hearing passive primes involving agent-patient and theme-experiencer verbs (Experiment 1), and theme-experiencer and experiencer-theme verbs (Experiment 2), than after corresponding active primes; moreover, the magnitude of priming was unaffected by verb type. However, a picture-sentence matching task (Experiment 3) replicated previous findings: Children performed more poorly on experiencer-theme sentences than agent-patient sentences. Our results suggest that children's acquisition of passive syntax is not delayed, and that semantic effects found in previous studies may instead be task-related. (Contains 8 tables and 3 figures.)