Phenotypic variation in human population may contain contributions from a number of different sex-associated genetic influences. These influences include maternal effects, the effects of sex-linked loci, and the effects of sex-limited autosomally linked loci. The families produced by MZ and DZ twins provide statistics which permit the detection and estimation of these effects. In particular, they provide statistics derived from various types of age-matched half-sibs and cousins in addition to those derived from the more usually studied full-sib or parent-offspring relationships. Specific models for genetic maternal effects, sex-linkage and sex-limitation are used to explore the use of extended twin design for the detection of and the discrimination between various sex-associated effects. The sample sizes required to detect maternal effects and sex-linkage were considered for some simple cases and it is concluded that comparison derived from the progeny of twins will often provide better tests for these effects than those derived from parent-offspring comparison.