We conducted field experiments in 1994 and 1995 to determine whether fleshy fruits consumed by North American, migrant passerines are antagonistic, complementary, or perfectly substitutable resources. Joint consumption of resources that are antagonistic results in reduced fitness, whereas joint consumption of those that are complementary resources results in increased fitness, than would be predicted from consumption of a linearly weighted sum of the two (or more) resources. Joint consumption of perfectly substitutable resources results in fitness equal to that predicted by consumption of the linearly weighted sum of the resources. Of ten pairs of fruit species tested, eight exhibited resource complementarity (sign test,P < 0.05); however, the strength of complementarity varied among the species pairs. In the two comparisons in which evidence for complementarity was not found, most fruit consumption appeared to have been accomplished by seed predators rather than true frugivores (house finches, Carpodacus mexicanus, in one case, and eastern chipmunk, Tamias striatus, in a second). In a replication of the second comparison, most consumption of one fruit species was due to a flock of starlings (Sturnus vulgaris), a species lacking sucrase activity (which may help explain the result). We suggest that resource complementarity is generally common among fruits, which may help explain why so few frugivores are found to specialize on only a single or even a small number of fruit species. Resource complementarity has important implications for the ecological and evolutionary interactions of avian frugivores and fruit-producing plants. The biochemical and physiological bases of complementarity are unknown, but could involve nutrients, secondary metabolites/toxins, or both.