The prostate specific antigen (PSA) level has become an important but imperfect means of detecting prostatic carcinoma. PSA index (serum PSA normalized to estimated gland volume) has been suggested to improve the performance characteristics of PSA alone. In an attempt to confirm this observation, we compared serum PSA alone to the PSA index in 218 men undergoing systematic random prostatic needle biopsy. Total gland PSA index as well as nontransition zone PSA index were calculated using several constants for the estimated contribution to the serum PSA from the transition zone. The Mann-Whitney nonparametric analysis was performed to account for differences in variances within the data set. For the patient population as a whole, all methods of testing were approximately equivalent in the ability to provide a statistically significant (p < 0.01) stratification between patients with benign and malignant biopsies. In patients with a serum PSA level of 4.1 to 10.0 ng./ml. none of the tests was able to distinguish those with carcinoma from those with a benign biopsy. In men with a normal prostate on digital rectal examination serum PSA was superior to other tests in predicting biopsy results. We conclude that PSA index does not enhance the ability of serum PSA alone to predict the presence of carcinoma.