Differentiation-inducing agents, such as retinoids and short-chain fatty acids, have an inhibitory effect on tumor cell proliferation and tumor growth in preclinical studies. Clinical trials involving these compounds as single agents have been suboptimal in terms of clinical benefit. Our study evaluated the combination of phenylbutyrate (PB) and 13-cis retinoic acid (CRA) as a differentiation and antiangiogenesis strategy for prostate cancer. On the basis of previous evidence, common signal transduction pathways and possible modulation of retinoid receptors and retinoid response elements by PB could be responsible for such activities. We assessed the effect of the combination of PB and CRA on human and rodent prostate carcinoma cell lines. The combination of PB and CRA inhibited cell proliferation and increased apoptosis in vitro in an additive fashion as compared with single agents (P < 0.014). Prostate tumor cells treated with both PB and CRA revealed an increased expression of a subtype of retinoic acid receptor (retinoic acid receptor-beta), suggesting a molecular mechanism for the biological additive effect. The combination of PB and CRA also inhibited prostate tumor growth in vivo (up to 82-92%) as compared with single agents (P < 0.025). Histological examination of tumor xenografts revealed decreased in vivo tumor cell proliferation, an increased apoptosis rate, and a reduced microvessel density in the animals treated with combined drugs, suggesting an antiangiogenesis effect of this combination. Thus, endothelial cell treatment with both PB and CRA resulted in reduced in vitro cell proliferation. In vivo testing using the Matrigel angiogenesis assay showed an additive inhibitory effect in the animals treated with a combination of PB + CRA (P < 0.004 versus single agents). In summary, this study showed an additive inhibitory effect of combination of differentiation agents PB and CRA on prostate tumor growth through a direct effect on both tumor and endothelial cells.