Lewis (1989) has suggested that the effect of so-called age incidence on season-of-birth data may be a sufficient explanation of the anomalies found in many samples of schizophrenic patients. Various predictions made by Lewis do not agree, however, with data available from some of the sources he quoted. Age incidence can be a source of error when a small age difference has a considerable effect on the likelihood of being included in a sample. Depending on methods of ascertainment, this is unusual in empirical studies.