Aerodynamic theory predicts that power output during flight should vary with body mass by an exponent of 1.56 when wing morphology remains constant (within an individual), and by an exponent of 1.19 when wing morphology changes with body mass (within a species or between species). I tested these predictions by estimating the power input during horizontal flight in three pregnant and two subadult Glossophaga soricina using a multivariate regression model. This analysis yielded power input during resting and flight as well as the energetic equivalent of change in body mass. A comparison of the estimated flight power for pregnant G. soricina, with published data on flight power of nonpregnant adults, revealed that energy turnover in flight is highest for pregnant G. soricina. Flight power of a 13-g pregnant G. soricina was even higher than that of a 16-g non-pregnant Glossophaga longirostris. A least-squares regression analysis yielded the following equations for the intraspecific scaling of flight power with body mass: power input during horizontal flight (Pf) = 24099 body mass (bm; kg)2.15 (r2 = 0.97) for the intra-individual allometry (pregnancy) and Pf = 113 bm(kg)0.95 (r2 = 0.99) for the inter-individual allometry (ontogeny). Both mass exponents are not significantly different from the predicted values for the scaling relationship of flight power within an individual (1.56) and within a species (1.19). This is the first measurement of power input during flight for subadult and pregnant bats.