The mechanism by which normal human prostate cells develop into prostatic carcinoma cells is not presently known. In the present study we have tested the hypothesis that specific prostatic carcinomas develop as a consequence of activation of a cellular gene(s) with transforming and tumorigenic potential. To test this possibility, high molecular weight DNA was extracted from the human prostatic carcinoma cell line, LNCaP, and cotransfected with a dominant acting neomycin resistance gene, pSV2-neo, into a subclone of Fischer rat embryo fibroblast (CREF) cells, CREF-Trans 6, and NIH-3T3 cells. Cells were selected for growth in G418 and pooled resistant colonies, which were morphologically normal, were injected subcutaneously into athymic nude mice. Tumors developed in several of the animals inoculated with LNCaP DNA-transfected CREF-Trans 6 cells and they were established in monolayer culture. In contrast, no tumors developed in nude mice injected with untransfected CREF-Trans 6 cells, pSV2-neo transfected CREF-Trans 6 cells or LNCaP plus pSV2-neo DNA-transfected NIH-3T3 cells. DNA from the first cycle tumor-derived CREF-Trans 6 cell lines, which were morphologically transformed in monolayer culture, was cotransfected with pSV2-neo a second time into CREF-Trans 6 cells and transfected cells, which were still morphologically normal, were injected into nude mice. Tumors developed in animals and they were again established in tissue culture. Secondary transfectants isolated from animals were morphologically transformed and grew with high efficiency in agar. Both primary and secondary LNCaP-transfected-nude mouse tumor derived-CREF-Trans 6 cells contained human repetitive (Alu) sequences. Although the pattern of Alu integration in the tumor derived CREF-Trans 6 cells were different for different tumors, both primary and secondary tumors contained a single apparently common-sized Alu fragment. The present study indicates that the human prostatic carcinoma cell line, LNCaP, contains a dominant-acting tumor-inducing oncogene which does not induce morphological transformation of CREF-Trans 6 or NIH-3T3 cells in monolayer culture. In addition, the CREF-Trans 6 cell line can detect this tumor-inducing gene function, whereas this activity is not observed in DNA-transfected NIH-3T3 cells.