This study used a relatively new method, syntactic priming, to investigate a persistent problem in language acquisition: children acquiring English as a first language comprehend passives with actional verbs earlier than passives with non-actional verbs (eg Maratsos, Fox, Becker & Chalkley, 1985). It has been argued therefore that children acquire the passive late or in a semantically constrained fashion. Although other studies have shown that children produce passives earlier, both in spontaneous speech and in elicited production experiments suggesting that the passive is acquired earlier than the comprehension results suggest, this research has not addressed the issue of semantic constraints. The present study used syntactic priming experiments to investigate this issue and, to anticipate, found reliable priming of passive targets but no effect of verb class. The nature of children’s early passives was further investigated in a follow-up experiment comparing priming from be-passives and get-passives. The results are discussed in relation to the language acquisition literature.