Vocal mimics that produce large repertoires of song types, such as in the Mimidae, have unique challenges discriminating songs of conspecifics from those of other mimids in areas where these species co-occur. We investigated cues used by Northern Mockingbirds (Mimus polyglottos) in discriminating their songs from songs of a sympatric mimid, the Brown Thrasher (Toxostoma rufum). We presented territorial mockingbirds with four playback treatments in which either mockingbird song types or thrasher song types had either a standardized mockingbird repetition pattern (5 repetitions) or a standardized thrasher pattern (2 repetitions). Four measures (time within 2 m of speaker, latency to approach, closest approach, and number of flights) were used to estimate a subject's response to each playback. Subjects responded significantly more strongly to mockingbird song types in a mockingbird repetition pattern than to thrasher song types in either repetition pattern. Responses to mockingbird song types in a thrasher repetition pattern elicited intermediate responses. Thus, mockingbirds can distinguish conspecific songs from Brown Thrasher songs based on song types alone regardless of their repetition pattern, although repetition pattern still appears to have a role in conspecific recognition. Brown Thrasher song includes a significantly broader frequency range than mockingbird song, which may allow direct discrimination. Our results suggest cues used by mimids in species discrimination are not necessarily the same as those used by human observers.