Treatment of Noble rats with separate silastic implants containing testosterone (T) and estradiol-17 beta (E2) for 16 weeks has previously been shown to induce multifocal epithelial dysplasia, a putative preneoplastic lesion, consistently in the dorsolateral prostate (DLP) but not in the ventral prostate (VP). We now studied effects of diethylstilbestrol (DES) substituted for E2 on these prostate lobes under the same conditions of exogenous androgen support. Three-week treatments with one 1-cm-long silastic implant of E2 or DES were approximately equipotent in changing target-organ weights and plasma prolactin. Accordingly, rats received for 16 weeks one 1-cm-long E2 or DES implant and two 2-cm-long T implants. In contrast to T + E2, T + DES induced widespread multifocal VP dysplasia and less or no DLP dysplasia. A serum-free explant-culture assay was used to determine uptake and metabolic disposition of 3H-labeled 5 alpha-dihydrotestosterone (DHT), T, and E2. Dysplastic VP explants incubated with 1.7 microM 1 beta-3H-labeled DHT and T accumulated more 3H-labeled steroid, metabolized 69 and 50% less substrate to terminal hydroxylated metabolites, and thereby formed and retained up to eight times as much estrogenic metabolite 5 alpha-androstane-3 beta,17 beta-diol (3 beta-androstanediol) and its lipoidal derivative than control VP. Experimental DLP explants did not form or retain more 3 beta-[3H]androstanediol than control DLP irrespective of treatments. Control VP metabolized [2-3H]E2 more actively to estrone than DLP. Dysplastic VP, however, metabolized one-half and accumulated five times as much E2 as VP and did not release more 3H as a marker of the 2,3-catechol estrogen pathway. These data suggest that differential target-tissue bioavailability of the estrogen component of the protracted dual-hormone stimulus determines in which prostate lobe dysplasia develops.