The hypothesis that puerperal affective psychosis (PAP) is genetically related to manic-depressive disorder was tested by comparing the morbidity risks for puerperal and non-puerperal affective disorders in the relatives of 17 PAP subjects and 20 parous manic-depressives (PMD) with no history of puerperal illness. The risk for affective disorder (mania, depression or suicide) and puerperal affective disorder was the same in the two groups of relatives and the test hypothesis was accepted, although the sample size was small. The frequencies of HLA-A, B and C locus antigens, nine blood group antigens and 10 red blood cell isoenzymes were not significantly different in the PAP and PMD subjects, showing that in this series genetic markers do not distinguish puerperal from non-puerperal affective psychoses.