A lthough dihydrotestosterone (DHT) is the principal androgen in the prostate, testosterone can also act as an androgen in this tissue. To determine the relative potencies of testosterone and DHT in preventing prostate regression, castrated rats were implanted for 4 d with varying doses of testosterone in the presence or absence of the 5alpha-reductase inhibitor finasteride. In the absence of finasteride, testosterone in the prostate is converted to DHT, creating an intraprostatic DHT dose response. In the presence of finasteride, this conversion is blocked, and an intraprostatic testosterone dose response is achieved. DHT was 2.4 times more potent than testosterone at maintaining normal prostate weight and duct lumen mass, a measure of epithelial cell function. The two androgens were equipotent at preventing DNA fragementation and expression of testosterone-repressed prostate message, two measures of apoptosis (cell death). The intraprostatic testosterone concentration that results from finasteride treatment in rats is sufficient to inhibit apoptosis but will not maintain normal epithelial cell activity. In conclusion, whereas DHT is more potent than testosterone at stimulating prostate epithelial cell function as measured by ductal mass, the two androgens are equipotent at preventing prostate cell death after castration. These results explain why finasteride causes prostate involution in the rat with minimal evidence of prostate cell death.