Allometric analysis was employed to compare linear dimensions of forelimb and hindlimb bones (humeri, radii, third and fifth metacarpals, third and fifth manual phalanges, femora, and tibiae) of 227 species of bats and 105 species of nonvolant mammals of varying degrees of phylogenetic affinity to bats. After accounting for body size, all forelimb bones are longer in bats than in nonvolant species, with the exception of humeri and radii of a few highly arboreal primates. Hindlimb bones are generally, but not uniformly, shorter in bats than in other mammals. For the humerus, radius, and metacarpals, midshaft diameters are greater in bats than in their comparably sized relatives. Proximal phalangeal midshaft diameters are statistically indistinguishable from those of other mammals, and distal phalanges show significantly reduced outer diameters. The pattern of relative reduction in wing bone diameters along the wing's proximodistal axis parallels the reduction in bone mineralization along the same axis, and a similar pattern of change in cortical thickness from the smallest wall thicknesses among mammals in the humerus and radius to the greatest wall thicknesses among mammals in the phalanges. The combination of altered cross-sectional geometry and mineralization appears significantly to reduce the mass moment of inertia of the bat wing relative to a theoretical condition in which elongated bones preserve primitive mammalian mineralization levels and patterns of scaling of long bone diameters. This intercorrelated suite of skeletal specializations may significantly reduce the inertial power of flight, contributing significant energetic savings to the total energy budgets of the only flying mammals.