Bovine viral diarrhea virus is a major cattle pathogen with a worldwide distribution. Animals may be infected with BVD virus transiently or persistently. Transient infection leads to protective immunity. Persistent infection is unique because it is associated with an immunotolerance that is specific to the infecting strain of BVD virus. Persistent infection results from viral invasion of fetuses between the second and fourth month of development. Such animals are of prime importance in the epidemiology of BVD because they shed large amounts of virus, and thus serve as a constant source of infection for non-immune animals. Infection of pregnant animals during the first two months of gestation may result in an increased rate of return to estrus. Animals infected in the period of five months to birth may abort or give birth to calves with malformations. The effects of BVD virus on fertility and gestation are well documented from experimental infection. However, much less is known of the extent of losses under field conditions. The main reason for this may be the multitude of other causes of increased return rates and gestation failures. In addition, the incidence of infection with BVD virus may vary over time and depends on management practices. In this study, we investigated the impact of BVD virus on gestation failures under field conditions in a large cattle-breeding area of Switzerland, where no specific measures to control BVD are in effect. Our approach consisted of relating seroconversions to BVD virus with the rate of return to estrus, abortion, and birth of calves with apparent malformations. These parameters of fertility were compared to those of animals immune to BVDV infection due to previous exposure to the virus and animals without seroconversion. Our data show that infection with BVD virus during the first 45 days of gestation did not influence the rate of return to estrus. By contrast, we observed a statistically significant increase in the abortion rate in mid-term gestation (Days 46 to 210) while no such effect was observed in the later stages of gestation. No clinically manifest malformations were observed in the offspring of animals that had seroconverted to BVD virus. In our study population the prevalence of BVDV antibody positive cattle varied only slightly between 78% and 80% over the period of observation. Our data showed that 7% (CI: 2.4-14%) of fetal deaths may be attributable to infection with BVD virus.