Recent public health research has consistently reported that self-identified multiracial adolescents tend to display more problem behaviors and psychological difficulties than monoracial adolescents. Relying on insights from qualitative analyses using small or clinical samples to interpret these empirical patterns, these studies implicitly assume a pejorative stance toward adolescents' multiracial self-identification. Building on the social psychological arguments underlying Park's and Stonequist's seminal discussions of the "marginal man," we derive hypotheses indicating that self-identified multiracial adolescents may show more psychological difficulties, but are also likely to have more active social interaction and participation than monoracial groups. We also incorporate later elaborations of the marginal man theory to develop alternative hypotheses regarding multiracial youth's school and behavioral outcomes. Based on a nationally representative sample of racially self-identified youth, the results suggest that patterns of multiracial-monoracial differences are generally consistent with the hypotheses derived closely from the marginal man theory or its subsequent elaborations. We examine the heterogeneities within these general patterns across different multiracial categories and discuss the implications of these findings. (Contains 1 figure, 6 tables and 19 notes.)