Two separate studies have shown a distortion in the sex ratio of first born children from HLA-DR compatible parents (Ober et al. 1985, Radvany et al. 1987). Kilpatrick (1987) was unable to confirm this distortion on an independent family data set. The question of the HLA-DR related sex ratio distortion still remained open. In the data set, "Provinces Françaises" families from 15 French provinces and Quebec were tested for a number of genetic markers including HLA-DR. In 1304 of these families, the HLA-DR results and family structure information were sufficient to allow the testing of this hypothesis. There were 2265 male and 2156 female children (overall sex ratio: 1.05); 1307 males and 1237 females were HLA-DR typed. In this group, the sex ratios are little different from those in the overall set, except for the firstborns which exhibit an apparent increase in the sex ratio. When dividing the sample between children fully HLA-DR compatible with their mother and those incompatible (i.e. having one antigen not present in the mother), the sex ratios in the two groups are little different whatever the birth order. This analysis has failed to observe any significant distortion in the sex ratio related to fetomaternal compatibility in agreement with the study of Kilpatrick. We conclude that, if such a distortion exists, it must be small.