OBJECTIVE To determine the significance of the minima of prostate-specific antigen (PSA) levels after orchidectomy in patients with advanced carcinoma of the prostate (stage C and D disease), to thereby discriminate patients who have a statistically significant difference in prognosis and thus individualize their adjuvant treatment. METHODS A retrospective analysis of 50 patients (mean age 75.8 years) with extracapsular prostate cancer (58% stage C, 42% stage D disease) was performed. PSA measurement and clinical assessment were continued at 3-monthly intervals until there was evidence of clinical progression. The minimum values of PSA were determined, the patients grouped at intervals of 1 ng/mL and the groups compared statistically on the basis of disease progression. RESULTS Serum PSA levels decreased in all patients to a minimum after 3-6 months. There was a statistically significant difference in the probability of, and time to, disease progression for PSA minima at 1 ng/mL and 10 ng/mL. Of the 13 patients in group 1 (with PSA minima < 1 ng/mL), 11 stayed in remission during a mean follow-up duration of 45.9 months. Of 25 patients in group 2 (with PSA minima between 1 and 10 ng/mL) 19 developed progression after a mean remission period of 16.7 months, while all 12 patients in group 3 (with PSA minima > 10 ng/mL) progressed after a mean remission period of 12.5 months. CONCLUSIONS Post-orchidectomy PSA minima are an excellent prognostic factor with significant predictive value. While patients in group 1 tend to have a favourable prognosis after orchidectomy alone, group 2 and 3 patients need adjuvant treatment as early as possible. Group 3 patients, with their high probability of early progression, also form a uniform cohort for studies concerning new therapeutic modalities. If the PSA minimum was included in further examinations, the results of treatment should improve.