The 17b-hydroxysteroid dehydrogenase type 3 (17b-HSD3) enzyme is a potential therapeutic target for hormone-dependent prostate cancer, as it is the key enzyme in the last step of testosterone (T) biosynthesis. A curcumin analog, H10, was optimized for inhibiting T production in LC540 cells that stably overexpressed 17b-HSD3 enzyme (LC540 [17b-HSD3]) (P < 0.01), without affecting progesterone (P) synthesis. H10 downregulated the production of T in the microsomal fraction of rat testes containing the 17b-HSD3 enzyme from 100 to 78.41 ± 7.41%, 51.86 ± 10.03%, and 45.14 ± 8.49% at doses of 10, 20, and 40 mM, respectively. There were no significant differences among the groups with respect to the protein expression levels of 17b-HSD3, 3bHSD1, CYP17a1, CYP11a1, and STAR, which participate in 17b-HSD3-mediated conversion of androgens to T (P > 0.05). This indicated that H10 only inhibited the enzymatic activity of 17b-HSD3 in vitro. Furthermore, H10 inhibited the adione-stimulated growth of xenografts established from LNCaP cells in nude mice in vivo. We conclude that H10 could serve as an effective inhibitor of 17b-HSD3, which in turn would inhibit the biosynthesis of androgens and progression of prostate cancer.