Syntactic priming effects, or the tendency to repeat syntactic structure across otherwise unrelated utterances, have been extensively investigated over the last two decades. Because such effects rely upon the language processor recognising a syntactic relationship between two utterances, they offer a versatile behavioural tool for investigating how syntactic structure is represented by the human cognitive system. In this paper, we review existing evidence to show that such effects can be truly syntactic in nature, and that they suggest syntactic representations that in many ways resemble strikingly those proposed by theoretical linguists on the basis of nonbehavioural measures. We suggest that there is a surprising dichotomy between the pervasiveness of such effects in contexts involving language production and their apparent elusiveness in language comprehension. We consider how syntactic priming research can cast light on issues of theoretical linguistic interest, including language use in special populations, and evaluate alternative explanations for the effects.