To study the efficacy and safety of bovine virus diarrhea virus (BVDV) vaccines there is need for a valid challenge model. We investigated whether sheep can be used in such a challenge model. We intranasally inoculated six groups (A-F) of seronegative sheep at day 49 of gestation with either of five antigenically different BVDV strains and one border disease virus strain. A seventh group (G) was housed for 10 days with a persistently infected calf and an eighth group (H) served as control. From each group half of the sheep were killed at 2 weeks, and half at 4 weeks after infection. For virus isolation five organs were collected from the sheep and seven from the fetuses. All sheep of groups A and H remained seronegative in the ELISA and in the serum neutralization test. At 2 and 4 weeks after infection virus was isolated from almost all fetal organs in six groups. In group A and in the control group no virus was isolated from the fetal organs. The virus distribution patterns in fetuses from sheep housed with the persistently infected calf or intranasally inoculated with the same strain were similar. We concluded that (i) antigenically different BVDV strains can induce congentital infection in sheep and that (ii) the consequences of a contact infection were similar to those after intranasal infection. In a second experiment we infected two groups of seronegative sheep with one of the strains used in the first experiment, before mating. A control group was left uninfected. The sheep were served and all sheep were challenged with antigenically homologous or heterologous BVDV at day 49 of pregnancy. Three weeks after challenge, sheep were killed and the procedure as in the first experiment was followed. None of the fetuses of the infected sheep were virus positive whereas all fetuses of the control sheep were virus positive. Hence, the immune response after BVDV infection protects fetuses against homologous and heterologous infection during pregnancy. Sheep may therefore be used in vaccination-challenge experiments to evaluate BVDV vaccine efficacy in preventing congenital infection.