Segregation analysis on five samples of families with Hunter's syndrome (158 cases overall) shows that the mutant aliele segregates in agreement with Mendelian expectations for an X linked recessive disease, but the proportion of sporadic cases is significantly lower than expected under mutationselection equilibrium. Heterogeneity among the samples is apparent, but it is caused entirely by a sample of Ashkenazi families, whose segregation pattern has previously been interpreted as supporting the hypothesis of prenatal selection in favour of the pathological aliele. Conversely, our joint analysis of the five samples by a maximum likelihood approach does not suggest segregation distortion. Possible reasons for the apparent lack of sporadic cases include the effect of ascertainment bias.