Many theories of verb learning propose that children can use information about a verb’s syntactic properties to make inferences about its meaning (syntactic bootstrapping, e.g. Gleitman 1990). On one bootstrapping hypothesis, children expect the number of participants in an event to match one-to-one the number of arguments of a verb describing that event—a heuristic called Participant-to-Argument Matching (PAM) (e.g. Naigles, 1990). Using this heuristic, a child who identifies that a verb occurs in a transitive or intransitive clause would be able to infer whether that verb describes an event readily perceived with 2 or 1 participants. However, previous studies have found inconclusive evidence that children use this heuristic with intransitive sentences (Noble, Rowland & Pine, 2011). In this talk I present a new test for PAM with intransitives. This study introduces a new method to test the fit between a sentence and a scene, and controls for methodological factors that may have influenced previous results. I’ll present some preliminary evidence that 19- to 22-month-olds do not consider an intransitive sentence to be a good fit for a 2-participant event, a result that would be consistent with a learning heuristic like PAM for intransitive sentences.