Estimation of the growth fraction of 153 prostatic carcinoma specimens employing Ki-67 immunostaining was undertaken and its relationship to various clinical parameters investigated. In prostate specimens, the percentage of tumour nuclei expressing Ki-67 antigen was measured and assigned a Ki-67 score. It was observed that high Ki-67 scores were associated with the poorly differentiated tumours, the correlation of this proliferation marker with histological grade was found to be significant (P less than 0.001). No relationship was observed between the Ki-67 score of the primary tumour with either the patient's age or with the primary tumor stage (T category). The metastatic status of the patient at diagnosis and the Ki-67 score of the tumour were correlated (P less than 0.05), higher Ki-67 scores being associated with M1 disease. Life-table analysis of 86 patients who subsequently received androgen withdrawal therapy, was undertaken with reference to the various Ki-67 scores of their primary tumors. A statistically significant difference in survival times was observed in patients whose Ki-67 values were less than 1% (P less than 0.0001) when compared to those patients whose tumours expressed 1% and over Ki-67 positivity, the former having longer survival times. When patients were subdivided according to their metastatic status and similar life-table analyses were carried out, no statistical difference was found between survival times and Ki-67 scores in M0 staged patients. In the M1 population of patients, however, those patients whose tumours were negative for Ki-67 expression had significantly longer survival times than those patients whose tumours exhibited positive Ki-67 staining (P less than 0.01). Comparing M1 staged patients whose prostate tumor cells exhibited less than 1% Ki-67 positive nuclei with M1 staged patients whose prostate tumour cells contained 1% and higher Ki-67 stained nuclei, a significantly longer survival time was found in the former group of patients (P approximately 0.0001).