At regional scales, the most important variables associated with diversity are latitudinally-based temperature and net primary productivity, although diversity is also influenced by habitat. We examined bird species richness, community density and community evenness in forests of eastern Connecticut to determine whether: 1) spatial and seasonal patterns exist in diversity, 2) energy explains the greatest proportion of variation in diversity parameters, 3) variation in habitat explains remaining diversity variance, and 4) seasonal shifts in diversity provide clues about how environmental variables shape communities. We sought to discover if our data supported predictions of the species–energy hypothesis. We used the variable circular plot technique to estimate bird populations and quantified the location, elevation, forest type, vegetation type, canopy cover, moisture regime, understory density and primary production for the study sites. We found that 1) summer richness and population densities are roughly equal in northeastern and southeastern Connecticut, whereas in winter both concentrate toward the coast, 2) variables linked with temperature explained much of the patterns in winter diversity, but energy-related variables showed little relationship to summer diversity, 3) the effect of habitat variables on diversity parameters predominated in summer, although their effect was weak, 4) contrary to theory, evenness increased from summer to winter, and 5) support for predictions of species–energy theory was primarily restricted to winter data. Although energy and habitat played a role in explaining community patterns, they left much of the variance in regional diversity unexplained, suggesting that a large stochastic component to diversity also may exist.