Rett syndrome is a neurodevelopmental disorder of unknown cause which affects girls almost exclusively. Apparently normal development in the first year of life is usually followed by loss of skills and the development of stereotypic hand movements. This study has used genetic epidemiological methods including a case control design to examine the evidence for aggregation of other disorders in families of girls with Rett syndrome. In one family there were two sisters with a condition consistent with Rett syndrome. Intellectual disability was not reported more commonly in case families (P = 0.46). However, "learning problems" were slightly commoner (P = 0.05) especially in the parental generation (P = 0.02) and these findings warrant further investigation. Mental illness and seizures were not reported at an increased prevalence. However, we would recommend the use of other strategies to collect information about psychiatric illness. Spinal curvature was reported more commonly in case families (P = 0.07) but no mechanism for clinical verification of this was included in the study. There was an apparent increase in bowel problems in the parents (P = 0.04). The major weaknesses of our study were our inability to validate any diagnosis clinically and the lack of power (due to the comparative rarity of the outcomes). The strengths are that we have been able to collect pedigree data on the families of a substantial proportion of a total population of girls with Rett syndrome and to collect comparative data from a control population. Our reported findings warrant further investigation in a larger study.