-The ecological and social interactions of the interspecifically aggressive Red-eyed (Vireo olivaceus) and Philadelphia (V. philadelphicus) vireos were studied during four breeding seasons in the Hubbard Brook Experimental Forest, New Hampshire. The two species here have extensively overlapping territories but occupy different vegetational strata during most of the breeding season. The aggressiveness of the Philadelphia Vireo towards the Red-eyed Vireo varies within a breeding season, being most frequent and intense during the periods of greatest vertical overlap. The aggressiveness of Red-eyed Vireos towards Philadelphia Vireos varies between years, being most frequent and intense in years of high Red-eyed Vireo population density. In one year no aggression was apparent between the two vireos. During some aggressive encounters, the smaller, subordinate Philadelphia Vireo used what appeared to be distraction displays when Red-eyed Vireos came close to their young. I hypothesize that resource competition may be the underlying cause of the aggression between these two ecologically similar species, largely because the most intense and frequent interspecific interactions occurred during periods of maximum overlap in foraging microhabitat. A comparison of these findings with those in another study illustrates the importance of forest height and vegetation structure in facilitating their coexistence. The ecological and evolutionary significance of interspecific aggression in birds has been the subject of controversy in recent years (Orians and Willson 1964, Cody 1969, 1974, Murray 1971, 1976). Support for the adaptiveness of interspecific aggression (cf. Cody 1974) has come from Rice's (1978a, b, c) study of the Red-eyed (Vireo olivaceus) and Philadelphia (V. philadelphicus) vireos where they occur together in the woodlands of northeastern Ontario. In this region the two species occupy exclusive non-overlapping territories, the smaller (12 g) Philadelphia Vireo successfully defending territories against its larger (17 g) congener. Rice (1978a) considered this horizontal separation to be a form of ecological segregation, because in nearly all other respects (habitat, use of vertical strata, foraging behavior) the two species were found to be essentially identical. In a study of the ecological relations of these same two species in a late successional forest in New England, I have observed that their territories overlap extensively and that interspecific aggression occurs at certain times during the nesting cycle. In this paper I examine the interaction of these two vireos at this site, compare the results to those of Rice (1978a, b, c) and consider reasons for the aggression between the two species. STUDY AREA AND METHODS This study was conducted in the 3,076-ha Hubbard Brook Experimental Forest, West Thornton, New Hampshire. In this late successional northern hardwoods forest the canopy layer (13-27 m) is dominated by sugar maple (Acer saccharum), American beech (Fagus grandifolia) and yellow birch (Betula alleghaniensis) with occasional white ash (Fraxinus americana). The sapling layer (2-12 m) is relatively sparse (Sherry 1979) and consists primarily of beech, sugar maple and some striped maple (A. pensylvanicum). The shrub layer (0-2 m) is a very dense and heterogeneous mixture of hobblebush (Viburnum alnifolium), striped maple, mountain maple (A. spicatum), sugar maple and beech. Most of my data were gathered on or near a 10-ha study plot at an elevation of 600 m. This plot is marked by a 50 m by 50 m grid, and has been the site of an intensive study of bird community dynamics and struc. ture since 1969 (Holmes and Sturges 1975, Holmes et al. 1979a). From late May through late July during 1975-1978 I visited this plot nearly every day for a total of over 200 days and 1,200 hours. Unless otherwise stated, all observations used in this paper were made on or within 100 m of the study plot. Other sections of the forest, especially those with Philadelphia Vireos, were also visited regularly each summer. To characterize the ecological and social relations of the two species, I measured their spacing, microhabitat use, habitat selection and reproductive success, each of which is treated separately below. Spacing. I mapped territories and estimated population densities of vireos on the study plot by methods similar to those of Holmes and Sturges (1975). Briefly, I made cumulative maps of all individuals heard, seen and/or captured in mist-nets and of their nest locations hen known. An average of 65% of the Red-eyed Vireos (REVs) on the 10-ha plot were color-banded by