Androgen-deprivation therapy has been the standard treatment for metastatic and locally advanced prostate cancer, but the majority of patients will progress to castration-resistant prostate cancer within 2-3 years. Unlike the case in breast cancer, no clinically validated biomarker has been used to predict the outcomes of androgen-deprivation therapy. To evaluate androgen-receptor splice variant-7 (AR-V7) detection in newly diagnosed advanced prostate cancer and describe the distinctive prognosis of this novel molecular subtype, this study retrospectively enrolled 168 newly diagnosed prostate cancer patients from 2003 to 2015 who received androgen-deprivation therapy. AR-V7 immunohistochemical staining was performed with a monoclonal antibody, and AR-V7 expression was determined using Immune-Reactive Score data. The association between nuclear AR-V7 expression and prognosis was determined. Multiple cause-specific Cox regression and stratified cumulative incidences were used to analyze the prognosis risk. Among the 168 patients, 32 (19%) were AR-V7-positive. Compared with the AR-V7-negative patients, the AR-V7-positive patients had significantly lower prostate-specific antigen response rates (P<0.001) to androgen-deprivation therapy and a much shorter time to castration-resistant prostate cancer (P<0.0001). In Kaplan-Meier analysis, the AR-V7-positive group showed markedly lower castration-resistant prostate cancer progression-free survival (P<0.0001) and much lower cancer-specific (P<0.0001) and overall survival (P<0.0001) both in all enrolled patients and in patients with metastases. AR-V7 positivity was a significant predictor of castration-resistant prostate cancer progression in multiple Cox regression (hazard ratio: 4.826; 95% CI: 2.960-7.869; P<0.001). AR-V7 immunohistochemical detection in newly diagnosed prostate cancer patients who are planning to receive androgen-deprivation therapy, especially those with metastases, is necessary and valuable for prognostic assessment. AR-V7-positive prostate cancer should be considered a novel prostate cancer subtype that should be distinguished upon initial biopsy. The main limitation of this study is its observational nature.