In a previous series of in vitro fertilization experiments with mice we found non-random combination of major histocompatibility complex (MHC) haplotypes in the very early embryos. Our results suggested that two selection mechanisms were operating: (i) the eggs selected specific sperm; and (ii) the second meiotic division in the eggs was influenced by the type of sperm that entered the egg. Furthermore, the proportion of MHC-heterozygous embryos varied over time, suggesting that non-random fertilization was dependent on an external factor that changed over time. As a higher frequency of heterozygous individuals correlated with an uncontrolled epidemic by MHV (mouse hepatitis virus), we suggested that MHV-infection might have influenced the outcome of fertilization. Here, we present an experiment that tests this hypothesis. We infected randomly chosen mice with MHV and sham-infected control mice five days before pairing. We recovered the two-cell embryos from the oviduct, cultured them until the blastocyst stage, and determined the genotype of each resulting blastocyst by polymerase chain reaction. We found the pattern that we expected from our previous experiments: virus-infected mice produced more MHC-heterozygous embryos than sham-infected ones. This suggests that parents are able to promote specific combinations of MHC-haplotypes during fertilization according to the presence or absence of a viral infection.