We examined the use of 87 palustrine and lacustrine wetlands by nongame water birds in central and eastern Maine using 3,527 h of observation (1,501 visits) made during April-August, 1977-85. Wetlands used by 15 species of water birds were distinguished from those not used, according to 20 habitat features. The species were the common loon (Gavia immer) , pied-billed grebe (Podilymbus podiceps), double-crested cormorant (Phalacrocorax auritus), American bittern (Botaurus lentiginosus), great blue heron (Ardea herodias), green-backed heron (Butorides striatus), osprey (Pandion haliaetus), bald eagle (Haliaeetus leucocephalus), northern harrier (Circus cyaneus), Virgima rail (Rallus limicola), sora (Porzana carolina), spotted sandpiper (Actitis macularia), common snipe (Gallinago gallinago), herring gull (Larus argentatus), and belted kingfisher (Ceryle alcyon). Predictive models of habitat use were developed for each species. Water birds were classified by similarity of habitats used, and species use was contrasted by wetland type. Smaller, isolated wetlands were used by fewer (P < 0.05) species than larger wetlands in complexes; many species had large area-requirements (pied-billed grebe, common loon, herring gull, double-crested cormorant, bald eagle) or preferred to use wetlands near other wetlands (common loon, herring gull, great blue heron, spotted sandpiper, osprey, bald eagle). Wetland area contributed more to overall variation in species richness on wetlands than wetland isolation, although on small wetlands (<3.6 ha) isolation was a better predictor of species richness than wetland area. Wetlands with intermediate amounts (33-66%) of emergent vegetation supported more species (P< 0.05) than closed (>66%) or open (<33%) wetlands. Low pH typified wetlands used by large-bodied piscivores (common loon, cormorant, osprey). Other water birds were associated with more densely vegetated, chemically buffered wetlands. Habitat features associated with wetland use by each waterbird species are reported, as are numerical responses of waterbird populations to wetland features and estimates of annual variation in habitat occupancy. Lacustrine wetlands supported a distinct, low diversity community of water birds, including most fish-eating species. Waterbird diversity at forested palustrine wetlands was intermediate between lacustrine communities and more species-rich assemblages at palustrine emergent and scrub-shrub wetlands. Regional variation in wetland characteristics and water bird use was associated with surficial geology, soils, and management practices. Management for nongame water birds in Maine should consider providing emergent and aquatic-bed vegetation with variable cover-to-water ratios, accommodating species-specific habitat needs, focusing on species of restricted distribution and low abundance, and maintaining wetland complexes. Bird use and habitat information from 87 wetlands and models of habitat selection for each species are provided in appendixes.