Abstract. We cloned a novel splicing variant for nuclear coactivator p120(α), designated as p120β and studied its function and expression in several human prostate diseases. Transfection assays demonstrated that p120β functions as a strong coactivator for androgen receptor (AR), but weakly for other nuclear receptors. GST-pull down assay showed that a glutamine-rich region of the p120 bound to the ligand-binding domain of AR. Interestingly, p120β mRNAs were expressed predominantly in the normal prostate, androgen-responsive prostate cancers and an androgen-sensitive prostate cancer cell line, LNCaP, but weakly in recurrent cancers and the androgen-insensitive prostate cancer cell lines PC3 and DU145. Furthermore, knockdown of p120α by siRNA abolished coactivator activity on thyroid hormone receptors (TR) and PPARγ, but did not affect that of ARs in PC3 cells. In addition, competitive assay with other nuclear receptors demonstrated that TR and PPARγ did not inhibit p120β-induced stimulation. These findings suggested that while p120α was essential for ligand-dependent stimulation of TRs and PPARγ, p120β acted as a coactivating protein predominantly for AR. PROSTATE cancer is the most common male malignancy in most Western countries, and the second leading cause of cancer death of men in the United States [1]. At the early stage of prostate cancer, the growth of prostate cancer cells is usually androgen-dependent and some tumor cells become androgen-independent in 6-18 months [2]. Androgen ablation through agonistic analogs of LHRH and antiandrogens, given either alone or in combination, can effectively treat prostate cancer [3]; however, hormone ablation therapy only causes temporary regression of prostate cancer. We have very few prognostic markers for prostate cancer recurrence because the molecular mechanism by which prostate cancer becomes androgen-independent is not clear. Therefore, the development of new diagnostic approaches to detect androgen-dependent or -independent tumor cells is important for improving the treatment and prediction of prognosis. Androgen receptor (AR) was the first factor highlighted as a cause of prostate cancer. Many investigations found mutation, amplification, overexpression, and a shorter CAG repeat sequence of the AR gene in androgen-independent prostate cancer [4][5][6][7][8]. The AR is a member of the steroid hormone superfamily of nuclear transcriptional factors. In the absence of ligand, AR is kept in a repressed state. In the presence of ligand, recruitment of coactivators such as SRC-1 [9]