Enzyme pro-drug suicide gene therapy has been hindered by inefficient viral delivery and gene transduction. To further explore the potential of this approach, we have developed AdIU1, a prostate-restricted replicative adenovirus (PRRA) armed with the herpes simplex virus thymidine kinase (HSV-TK). In our previous Ad-OC-TK/ACV phase I clinical trial, we demonstrated safety and proof of principle with a tissue-specific promoter-based TK/pro-drug therapy using a replication-defective adenovirus for the treatment of prostate cancer metastases. In this study, we aimed to inhibit the growth of androgen-independent (AI), PSA/PSMA-positive prostate cancer cells by AdIU1. In vitro the viability of an AI- PSA/PSMA-expressing prostate cancer cell line, CWR22rv, was significantly inhibited by treatment with AdIU1 plus GCV (10 microg ml(-1)), compared with AdIU1 treatment alone and also cytotoxicity was observed following treatment with AdIU1 plus GCV only in PSA/PSMA-positive CWR22rv and C4-2 cells, but not in the PSA/PSMA-negative cell line, DU-145. In vivo assessment of AdIU1 plus GCV treatment revealed a stronger therapeutic effect against CWR22rv tumors in nude mice than treatment with AdIU1 alone, AdE4PSESE1a alone or in combination with GCV. Our results demonstrate the therapeutic potential of specific-oncolysis and suicide gene therapy for AI-PSA/PSMA-positive prostate cancer gene therapy.