Background. We assess claims that the documented rise in psychiatric morbidity during the nineteenth and twentieth centuries was associated with an increasing incidence of schizophrenia.

Method. Cross-sectional epidemiological comparison of the incidence of schizophrenia in one urban, industrialized community at three points over more than 100 years using new data from 1881–1902 and two pre-existing datasets, 1978–1980 and 1992–1994. For 1881–1902, 34 cases of schizophrenia were obtained through retrospective diagnosis, using Research Diagnostic Criteria, of a random 14·5% sample of first admissions to Nottingham Asylum (n=330). Inter-rater reliability and leakage analyses were performed. The administrative incidence for all three studies was directly standardized against 1991 census data. Local statistics on total psychiatric morbidity in Nottingham were taken from the asylum superintendent's register and recent data from the Office of National Statistics.

Results. Official local and national rates of total psychiatric morbidity increased exponentially. There was no significant change in the incidence of schizophrenia over the 114-year period 1881–1994.

Conclusions. The rise in both local and national official statistics of psychiatric morbidity 1881–1994 was not associated with a significant increase in the incidence of schizophrenia. Stability in the epidemiology of schizophrenia at a geographical level is found despite important demographic changes.