Long-term effects of disease on wild animal population demography is not well documented. We studied a gray wolf (Canis lupus) population in a 2,060km2 area of Minnesota for 15 years to determine its response to canine parvovirus (CPV). The CPV had little effect (P gt 0.05) on wolf population size while epizootic during 1979-83. However, after CPV became enzootic, percentage of pups captured during summer-fall 1984-93 and changes in subsequent winter wolf numbers were each inversely related to the serological prevalence of CPV in wolves captured during July-November (r2 = 0.39 and 0.72, P = 0.05 and lt 0.01, respectively). The CPV antibody prevalence in adult wolves increased to 87% in 1993 (r2 = 0.28, P = 0.05). However, because population level remained stable, CPV-induced mortality appeared to compensate for other mortality factors such as starvation. We -predict that the winter wolf population will decline when CPV prevalence in adults consistently exceeds 76%. The CPV may become important in limiting wolf populations.