If mothers are carriers of a recessive male sex-linked disorder, their children may include females, healthy males and affected males. The composition of the possible sibships follows a multinomial distribution. If the mothers are diagnosed as carriers independently of their offspring, then the distributions are not truncated, but if they are classified as carriers after the birth of at least one affected son, the distributions are truncated. In order to avoid ascertainment bias, the statistical analyses of the corresponding data must differ. We present procedures for estimation in both the untruncated and the truncated case. The formulae obtained are applied to family data for X-chromosomal recessive retinoschisis (RS) from the region of Satakunta in southwestern Finland. Earlier studies indicated that, among the offspring of RS carriers, the sex ratio was increased. In our model, therefore, we include both the proportion of males and the proportion of affected individuals among the males. The sex ratio was significantly increased (131.0), but the proportion of affected males was slightly lower than expected (42.5%). RS seems to be the only known genetic disorder in which the offspring of carriers of the mutation include a statistically significant surplus of males.