Endocrine therapy, which consists of orchiectomy followed by administration of large doses of estrogen, then a reduced amount of estrogen, has been applied as the main treatment for stage D2 prostatic cancer. Alternatively, anti-androgen is used for elderly patients or those with cardiovascular disorders. Survival rate with endocrine therapy at 5 and 10 years was 35% and 16%, respectively. Therefore, in Japan, a better survival is shown than that reported in western countries using much smaller doses of estrogen. Most of the side effects caused by estrogen are not serious. Side effects caused by anti-androgen are few except for loss of libido. At the start of treatment, more than 80% of patients showed a response, but gradually relapse occurred and only 20% were well controlled 5 years after the start. Factors influencing the survival were pathological grade, response to endocrine therapy judged by the level of prostatic acid phosphatase 4 weeks after the start, and R1881 (methyltrienolone)-binding protein observed histochemically. The latter protein was also correlated with the grade and response to endocrine therapy. Relapse after endocrine therapy might be attributable to adaptation or mutation progressing to androgen-independent cells. Using SC 115, an androgen-dependent mouse tumor, these two types of relapse were demonstrated. Gradual progression to undifferentiated cancer was noticed between pretreatment biopsy and autopsy. Relapse in human prostatic cancer may thus be partly due to genetic change to a resistant clone.