Reproductive technology, especially the diagnosis of pregnancy by radioimmunoassay of fecal steroid metabolites, is an important component of captive propagation, but its role in our understanding of ecological interactions and in situ biological restoration has been more limited. Where large herbivores have been 'released' from predation by the extirpation of carnivores, controversy often exists about possible detrimental effects at the ecosystem level A related concern is that the reestablishment of large carnivores may decrease the availability of prey populations for human subsistence. We suggest that pregnancy assays can be a valuable tool to help distinguish between the roles of predation versus food-imposed limitations on population size and their effects on juvenile recruitment in wild species. We explored this issue through analyses of fecal progestagen concentration (FPC) levels to document pregnancy in moose (Alces alces) in the southern Greater Yellowstone Ecosystem, a site where wolves (Canis lupus) and grizzly bears (Ursus arctos) are recolonizing former habitats after an absence of more than 60 years. Pregnancy was clearly discernible (mean FPC for pregnant and nonpregnant females, respectively: 10.60 vs. 2.57 ??g/g; p < 0.0001). Among the potential confounding variables that need to be considered if FPC is applied to ecological and demographic questions are whether baseline values are affected by handling, whether neonate survival has been assessed, and whether sampling efforts are directed at both pregnant and nonpregnant animals. With these issues accounted for, the local moose population experienced juvenile survival rates among the highest in North America. Pregnancy rates, however, dropped from 90% in 1966 to about 75% today, rendering them in the lowest fifteenth percentile among moose populations in North America. Our findings suggest that a relatively low frequency of juvenile moose is not the likely result of predation, and they illustrate how endocrinology can be applied to issues involving reproductive events within an ecological context. They also affirm that noninvasive and generally inexpensive endocrinological procedures will be applicable to understanding interactions between recolonizing predators and prey, an issue that will continue to arise because of global restoration efforts, and to the study of rare ungulates in remote systems where data on reproductive events are difficult to obtain.