Gametes, somatic cells and materials of animal origin in media are potential sources for introducing bovine viral diarrhea virus (BVDV) into systems for production of IVF bovine embryos. Further, the efficacy of washing and trypsin treatment for removal of BVDV from IVF embryos is questionable. Washing and trypsin treatments recommended by the International Embryo Transfer Society for in vivo-derived embryos were applied to in vitro-derived, virus-exposed, bovine embryos in this side-by-side comparison of treatments. Embryos for the study were produced in a virus-free system in which follicular oocytes were matured and fertilized in vitro and presumptive zygotes were co-cultured with bovine uterine tubal cells for 7 d. A total of 18 trials was performed, 9 using a noncytopathic BVDV and 9 using a cytopathic BVDV. In each trial, 4 equal groups of 10 or less, zona pellucida-intact embryos/ova were assembled, including 2 groups of morulae and blastocysts (M/B) and 2 groups of nonfertile or degenerated ova (NFD). Each group was prewashed and exposed to 10(4) to 10(6) TCID50/mL of either noncytopathic (SD-1) or cytopathic (NADL) BVDV for 2 h. Following in vitro viral exposure, one group of M/B and one group of NFD were washed. The other groups of M/B and NFD were trypsin-treated, Both treatments were consistent with IETS guidelines. After in vitro exposure to noncytopathic BVDV and washing, viral assays of 100 % (9/9) and 78 % (7/9) of the groups of M/B and NFD ova, respectively, were positive. After in vitro exposure to cytopathic BVDV and washing, viral assay of 33 % (3/9) of the groups of both M/B and NFD ova were positive. After in vitro exposure to noncytopathic BVDV and trypsin treatment, viral assay of 44 % (4/9) of groups of M/B and 67 % (6/9) of groups of NFD ova were positive. Finally, after in vitro exposure to cytopathic BVDV and trypsin treatment, viral assay of 22 % (2/9) of the groups of M/B and 44 % (4/9) of the groups of NFD ova were positive. Contingency table analysis, in which data was stratified by embryo type and virus biotype, was used to compare results. While a difference existed between results of the 2 treatments of groups of M/B within the noncytopathic biotype (P = 0.01, Mantel Haenszel Chi-square), no difference was observed between comparison of treatment between all groups in both biotypes (P > 0.05). (C) 1998 by Elsevier Science Inc.