The object of this study was to determine whether interspecific competition modified local geographic distribution, after taking into account the effect of habitat structure. The tendencies for 14 passerine birds to have positive or negative associations were examined, using 7861 sample points in seven native forests on the islands of Hawaii, Maui, and Kauai. All birds were at least partly insectivorous and were fairly common in forested areas, although some fed chiefly on nectar or fruit. Species-pairs were classified as primary or secondary potential competitors based on general dietary similarity. To evaluate the association between species and to account for the effect of individual species habitat preferences, partial correlations were computed for each species-pair in a study area from the simple correlations between the species and 26 habitat variables plus two quadratic terms to represent nonlinearity. The partial correlations represented a short-term ('instantaneous') assessment of the strength of competitive interactions, and did not reflect the accumulation of competitive displacement through time. Of 170 partial correlations in the analysis, only 10 indicated significant negative association. The general pattern was of positive association (76 significantly positive partials), which probably resulted from flocking and from attraction of birds to areas of resource superabundance. Two species showed consistent patterns of negative partial correlations over several adjacent study areas, the Japanese White-eye/Iiwi in montane Hawaii, and the Japanese White-eye/Elepaio in windward Hawaii; both patterns could be reasonably attributed to direct competition. Species-pairs were grouped by the native or exotic status of the component species. Native/exotic pairs had a significantly greater proportion of negative partial correlations (37%) than either native/native pairs (8%) or exotic/exotic pairs (0%). This pattern was consistent across the seven study areas and appeared to reflect the occurrence of interspecific competition along a broad and diffuse ecological 'front' between a co-evolved native avifauna and recently introduced exotic species. The role of competition in the pattern was corroborated by the significantly higher proportion of negative partial correlations among species-pairs of primary potential competitors than among those of secondary potential competitors. Our results suggested that 47% of the primary potential competitors among native/exotic species-pairs may experience at least small depressions in local population density due to competition. Although the negative correlations were for the most part small (average negative r = 0.06), one species could eventually replace another as spatial displacement accumulated through time. The Japanese White-eye appeared to have a principal role in native/exotic interactions, with 62% of the partial correlations between it and native primary potential competitor species being negative. Noteworthy implications were that (1) it was important to account for the habitat responses of individual species when studying the role of interspecific competition in modifying small-scale geographic distribution; (2) competition was frequently sporadic in its geographic occurrence and in the species affected, thus supporting Wiens' (1977) theory of competition; and (3) as a consequence, the role of interspecific competition in modifying distribution may be difficult to detect statistically with small data sets.