SUMMARY In order to study the cumulative effect of mouse histocompatibility antigens, donor-host combinations were utilized which differed at 2, 3, 4, and multiple histocompatibility loci. Double difference pairs involving the Y-linked and one of several autosomal loci were produced by grafting from a male of one strain to a female of its congenic resistant partner strain. Double difference pairs were also produced by grafting between strains each having a single distinct autosomal difference from a common congenic resistant partner. Triple-difference pairs were produced by using a male donor and a female recipient, each having a different single autosomal histocompatibility difference from a common congenic resistant partner. A quadruple-difference pair was produced by using a male donor and a female recipient, one having a double autosomal histocompatibility difference from a common congenic resistant partner. The inbred strains C57BL/10 and 129 were utilized to study multiple gene differences. Grafts exchanged between such double, triple, quadruple, and multiple-difference pairs showed a cumulative effect if the discrepancies between the median survival times of the constituent pairs were not large. Conversely no cumulative effect was shown if the discrepancies between the median survival times of the constituent pairs were large. It made no difference if the pairs differed at non-H-2 loci or at a non-H-2 locus and the H-2 histocompatibility locus. In the absence of strong histoincompatibilities, otherwise weak and insignificant immunogens became signifi-