Androgen insensitivity syndrome encompasses a wide range of phenotypes, which are caused by numerous different mutations in the AR gene. Detailed information on the genotype/phenotype relationship in androgen insensitivity syndrome is important for sex assignment, treatment of androgen insensitivity syndrome patients, genetic counseling of their families, and insight into the functional domains of the AR. The commonly accepted concept of dependence on fetal androgens of the development of Wolffian ducts was studied in complete androgen insensitivity syndrome (CAIS) patients. In a nationwide survey in The Netherlands, all cases (n = 49) with the presumptive diagnosis androgen insensitivity syndrome known to pediatric endocrinologists and clinical geneticists were studied. After studying the clinical phenotype, mutation analysis and functional analysis of mutant receptors were performed using genital skin fibroblasts and in vitro expression studies. Here we report the findings in families with multiple affected cases. Fifty-nine percent of androgen insensitivity syndrome patients had other affected relatives. A total of 17 families were studied, seven families with CAIS (18 patients), nine families with partial androgen insensitivity (24 patients), and one family with female prepubertal phenotypes (two patients). No phenotypic variation was observed in families with CAIS. However, phenotypic variation was observed in one-third of families with partial androgen insensitivity resulting in different sex of rearing and differences in requirement of reconstructive surgery. Intrafamilial phenotypic variation was observed for mutations R846H, M771I, and deletion of amino acid N682. Four newly identified mutations were found. Follow-up in families with different AR gene mutations provided information on residual androgen action in vivo and the development of the prepubertal and adult phenotype. Patients with a functional complete defective AR had some pubic hair, Tanner stage P2, and vestigial Wolffian duct derivatives despite absence of AR expression. Vaginal length was functional in most but not all CAIS patients. The minimal incidence of androgen insensitivity syndrome in The Netherlands, based on patients with molecular proof of the diagnosis is 1:99,000. Phenotypic variation was absent in families with CAIS, but distinct phenotypic variation was observed relatively frequent in families with partial androgen insensitivity. Molecular observations suggest that phenotypic variation had different etiologies among these families. Sex assignment of patients with partial androgen insensitivity cannot be based on a specific identified AR gene mutation because distinct phenotypic variation in partial androgen insensitivity families is relatively frequent. In genetic counseling of partial androgen insensitivity families, this frequent occurrence of variable expression resulting in differences in sex of rearing and/or requirement of reconstructive surgery is important information. During puberty or normal dose androgen therapy, no or only minimal virilization may occur even in patients with significant (but still deficient) prenatal virilization. Wolffian duct remnants remain detectable but differentiation does not occur in the absence of a functional AR. In many CAIS patients, surgical elongation of the vagina is not indicated.