摘要：Purpose We examined the clinical significance of long-term serum testosterone monitoring to predict the prognosis of patients with prostate cancer treated with combined androgen blockade. Materials and Methods We retrospectively analyzed the records of 225 patients who underwent combined androgen blockade as first line therapy for prostate cancer. The prognostic values of testosterone and other clinical factors were evaluated with respect to prostate specific antigen progression-free and overall survival. Results Median patient age was 73.0 years, median prostate specific antigen was 42.6 ng/ml and median followup was 45.8 months. No variable associated with testosterone was predictive of progression-free survival. With regard to overall survival on univariate analysis nadir testosterone less than 16 ng/dl (p = 0.0190), less than 20 ng/dl (p = 0.0020) and less than 32 ng/dl (p = 0.0146) were significant together with other clinical factors. In contrast, nadir testosterone less than 8 and less than 12 ng/dl were not significant. Multivariate analysis showed that nadir testosterone less than 20 ng/dl was the significant prognostic factor (p = 0.0048). In addition, time to nadir testosterone was about 1 year (11.3 months). Patients were divided into rapid and slow types based on time to testosterone less than 20 ng/dl before and after 6 months, respectively. No significant difference in overall survival was observed between the 2 types. The current results suggest that the critical factor for prognosis was not a rapid decrease but whether nadir testosterone achieved a level of less than 20 ng/dl. Conclusions Nadir testosterone 20 ng/dl was the most significant cutoff level for overall survival in Japanese patients with prostate cancer treated with combined androgen blockade.