OBJECTIVE: To investigate recent trends in prostate cancer incidence and mortality, with particular reference to changes in diagnostic techniques and treatment. METHODS: The Thames Cancer Registry was used to identify all men, resident in SE England, diagnosed with prostate cancer between 1990 and 1999. Information regarding prostate cancer mortality was obtained from the Office of National Statistics. Other data sources were used to ascertain the number of transurethral resections of the prostate (TURP) and open prostatectomies performed in SE England, and the number of prescriptions issued for the treatment of benign prostatic hyperplasia (BPH). RESULTS: There was a steady increase in the age-standardised incidence of prostate cancer from 1990, which then began to plateau in 1996. The increase was entirely restricted to localised tumours; non-localised tumours showed a slight downward trend over this period. Age-standardised mortality rates have remained constant, with a slight fall in 1997 corresponding to the decline in incidence rates. Medical treatment for BPH has increased, with a corresponding reduction in the number of TURPs. CONCLUSION: The change in occurrence of prostate cancer is entirely due to changes in the incidence of localised cases. Incidence of non-localised cases and mortality remained almost constant. The increasing tendency in incidence of localised prostate cancer is likely to be principally due to increased detection, through increased use of prostate-specific antigen (PSA) testing followed by radical resections of the prostate. The aggregate effect of PSA testing and medical treatment of BPH is a stabilisation in the incidence level of localised cases in recent years.