Objective To determine the incidence and natural history of stage T1a‐T1b prostate cancer in patients undergoing surgery for benign prostatic hypertrophy (BPH), and thus evaluate the effect that recent medical and ‘minimally invasive’ treatments (which provide no prostate sample for pathological examination) might have on the percentage of patients with unsuspected prostate cancer. Patients and methods A series of 1648 patients undergoing surgery for BPH over a 13‐year period were reviewed retrospectively; the period overlapped the introduction of serum prostate specific antigen (PSA) as a detection method. Results Stage T1 prostate cancer was found in 182 patients (11%), comprising 126 (11%) of 1199 transurethral resections and 56 (12%) of 449 open enucleations. The introduction of systematic PSA assays gradually reduced the mean incidence of T1 cancer from 23% to 7%, with a greater effect on T1b (from 15% to 2%), while the incidence of T1a remained nearly constant (±5%). The pathological features of surgical specimens from 43 radical prostatectomies undertaken for T1 tumours were reviewed. Locally advanced disease (stage ≥pT3) was apparent in 13% of T1a and 28% of T1b tumours. Amongst the patients electing for surveillance, only 8% of those with T1a progressed within 30–97 months of follow‐up (mean progression time 73 months), whereas 29% of those with stage T1b progressed within 36 months of follow‐up (mean progression time 17 months). Conclusion These results show that the use of the PSA assay has decreased but not suppressed the incidence of pT1 prostate cancer, with a greater effect on those tumours at higher risk of progression (T1b). This suggests that the detection of prostate cancer based on PSA and transrectal ultrasonography is appropriate for screening patients and is sufficiently accurate that treatments for BPH that provide no pathological materials can be applied safely.