-Populations of migrant warblers were correlated with populations of lepidopterous larvae in spring in a heavily forested area of southern Illinois. The birds appeared to feed entirely on these larvae, mostly taken from oaks. Estimates of gross calorie intake indicated a positive energy balance for warblers in an area with larval biomass of 0.70 g per m3, and a negative balance in an area with biomass of 0.02 g per m3. Warblers chose either the smaller size classes (under 15 mm) of larvae, or the smaller species (especially leafrollers) out of proportion to their availability, but the birds' positive energy balance was dependent upon an irruption of geometrids. Lepidopterous larvae constituted 7598% of the available foliage invertebrates, and other groups were insufficient alone to support migrant populations. Larvae were about 80% water, and warblers ingested 1.2-1.7 times their own weight in larvae per day in order to attain gross caloric intake of 14.8-19.0 kcal per bird-day. Warblers in an area with insufficient food resources ingested 7.2 kcal/day, and their behavior was consistent with a goal of energy conservation. Among the problems faced by migrants after a flight is that of finding adequate food supplies to continue the migration, particularly as the landing area may be completely unknown to the birds. In Illinois the only measurements of migrant populations in forest habitat have been those of Twomey (1945) and Calef (1953) who found four to eight times more birds in elmmaple forest at the peak of migration in May than were present in June. We wondered how the forest accommodated large numbers of migrants, and how much forest is required by arboreal migrants. Kendeigh's (1979) study of invertebrates in three east-central Illinois study areas showed the seasonal population patterns of various groups of invertebrates. There was a close similarity between the timing of emergence of lepidopterous larvae and the seasonal timing of bird migration in this region. Stimulated by these papers and the question of how migrant populations respond to habitat availability, we began, in 1979, concurrent studies of bird populations and their potential food resources in (1) an area where forest tracts were extensive, and (2) an area where forest was much more restricted. Our primary goal was to consider the availability of food supplies to migrant warblers. Our interest was not in the influence of bird populations on insect populations, but the reverse. In either case, it is important to know about rates of predation. When the opportunity arose in May 1981 to study food consumption by warblers on invertebrates in forest-edge habitat at two widely divergent population levels, we interrupted our principal study to make the observations reported her .