We conducted field and laboratory experiments to evaluate whether treating pregnant bighorn ewes with a combination of an experimental Pasteurella trehalosi and Mannheimia haemolytica (formerly P. haemolytica) vaccine and a commercially-available bovine P. multocida and M. haemolytica vaccine would increase lamb survival following a pneumonia epidemic. Three free-ranging bighorn herds affected by pasteurellosis outbreaks between November 1995 and June 1996 were included in the field experiment. Post-epidemic lamb survival was low in all three herds in 1996, with November lamb:ewe ratios of < or = 8:100. In March 1997, thirty-six ewes (12/herd) were captured and radiocollared. Half of the ewes captured in each herd were randomly selected to receive both vaccines; the other half were injected with 0.9% saline solution as controls. Lambs born to radiocollared ewes were observed two or more times per week and were considered to have survived if they were alive in October 1997, about 6 mo after birth. Lamb survival differed among herds (range 22% to 100%), and survival of lambs born to vaccinated ewes was lower (P = 0.08) than survival of lambs born to unvaccinated ewes. Bronchopneumonia (pasteurellosis) was the dominant cause of mortality among lambs examined. We concurrently evaluated vaccine effects on survival of lambs born to seven captive ewes removed from the wild during the 1995-96 epidemic. Antibody titers were high in captive ewes prior to vaccination, and vaccines failed to enhance antibody titers in treated captive ewes. None of the captive-born lambs survived. These data suggest that, using existing technology, vaccinating bighorn ewes following pneumonia epidemics has little chance of increasing neonatal survival and population recovery.