A maximum-likelihood scoring technique for analysis of pedigree data allows for the concurrent estimation of random and of fixed effects in a quantitative trait. We included both types of effects in genetic models, to study the sources of variation in finger ridge count in 54 large families affected with the fragile-X disorder. The fixed effects were represented by fragile X and sex, and the random effects by environmental and genetic variance. We found a significant effect of fragile X in the mean of the finger ridge count on the thumb (finger 1) and index finger (finger 2), which had the lowest heritability and a negligible nonadditive component of genetic variance. This was in contrast with ridge counts on fingers 3 and 4, which showed little fragile-X effect, but high heritability and a significant nonadditive component. A contrast in genetic properties for ridge counts on fingers 1 and 2, compared with these counts on the remaining three fingers, may be relevant to increased selection pressures on functions of the thumb and of the index finger in evolution of modern man. We have also demonstrated an important effect of fragile X in increasing the additive variance in covariance, especially between male pairs. These findings suggest that the effect of the fragile-X genotype in finger ridge count is additive and superimposed on the normal hereditary variations in this trait.