OBJECTIVE The "estrogen hypothesis" posits that this hormone serves as a protective factor in the development of schizophrenia. If true, then it is expected that the earlier the age of menarche, the later the onset of schizophrenia (as has been reported by some investigators). This study attempts to replicate this relationship in a sample of women with schizophrenia and schizoaffective disorder. METHODS Self-report menarche age, clinical status, and onset of disorder were collected in a sample of 68 women (55 diagnosed with schizophrenia and 13 with schizoaffective disorder). RESULTS Menarche age and schizophrenia onset were not negatively correlated as would be predicted by the estrogen hypothesis. Two clinical measures, however, did correlate with age of menarche as predicted. Higher negative symptom scores (total SANS) and greater functional impairment (lower GAF) were reported in subjects who reported a later age at menarche. CONCLUSIONS This study suggests that an earlier age at menarche might predict improved clinical outcome after schizophrenia onset (in support of the estrogen hypothesis). Our data, however, do not support Cohen et al.'s findings regarding the relationship between age at menarche, and the timing of the onset of the disorder. Further investigations regarding the relationship between estrogen and schizophrenia development in women are needed. It is suggested that other developmental factors, both biological and psychosocial, might play a crucial role in both the age at onset and the outcome of the disorder in women.