The incidence of congenital toxoplasmosis was determined by an ELISA in the litters of BALB/c mice which had been infected 8 weeks before mating, on day 12 of pregnancy, or on both these occasions. Of those mice given the infection for the first time on day 12 of pregnancy, 5 out of 6 gave birth to infected litters with approximately 50% of the individuals in each litter being infected. BALB/c mice which had been infected 8 weeks before mating did not give birth to infected litters, even if they were reinfected on day 12 of pregnancy. Following infection BALB/c mice were found to harbour significantly fewer tissue cysts than the congenic H-2 derivative BALB/K strain. However, chronically infected BALB/K mice also failed to produce infected litters, indicating that tissue cyst burden in the dam did not influence congenital infection at least on the BALB background. This study demonstrates that BALB/c dams chronically infected with Toxoplasma gondii, have immunity capable of protecting their embryos from congenital infection, even if the dams are reinfected during pregnancy. Our results demonstrate that the BALB/c mouse can be used as a model of human or ovine congenital T. gondii infection suitable for testing putative vaccines.