AIMS: To assess the possible reasons for error in the diagnosis of prostatic cancer with available follow-up data. METHOD AND RESULTS: A cohort of 1791 cases of prostatic cancer diagnosed in the UK between 1990 and 1996 was examined. All cases were clinically localized at presentation, treated by non-curative methods and detailed follow-up was available. A panel of genitourinary pathologists reviewed the pathology of all cases. One hundred and thirty-three (7.5%) of cases were reassigned to a non-malignant diagnosis. Where possible, reasons for the initial diagnosis were given. These included severe atrophy, inflammatory induced atypia, sclerosing adenosis, atypical adenomatous hyperplasia and basal cell hyperplasia. Follow-up of these patients showed an extremely low death rate from prostatic cancer: lower than that for the Gleason combined score of five or less tumours diagnosed in this series. CONCLUSIONS: Many morphological entities potentially mimic prostatic cancer and may be responsible for misdiagnosis in routine specimens. Continuing education in prostatic morphology and immunohistochemistry may have helped reduce this error rate.