To evaluate the long-term efficacy and toxicity of a hypofractionated (55 Gy in 20 fractions within 4 weeks) vs. a conventionally fractionated (64 Gy in 32 fractions within 6.5 weeks) dose schedule for radiotherapy (RT) for localized carcinoma of the prostate. A total of 217 patients were randomized to either the hypofractionated (n=108) or the conventional (n=109) dose schedule. Most patients (n=156) underwent RT planning and RT using a two-dimensional computed tomography method. Efficacy using the clinical, radiologic, and prostate-specific antigen data in each patient was evaluated before RT and at predetermined intervals after RT until death. Gastrointestinal and genitourinary toxicity using the modified Late Effect in Normal Tissue-Subjective Objective Management Analytic (LENT-SOMA) scales was also evaluated before and at intervals after RT to 60 months. The whole group has now been followed for a median of 90 months (range, 3-138). Of the 217 patients, 85 developed biochemical relapse (nadir prostate-specific antigen level+2 μg/L), 36 in the hypofractionated and 49 in the conventional group. The biochemical relapse-free, but not overall, survival at 90 months was significantly better with the hypofractionated (53%) than with the conventional (34%) schedule. Gastrointestinal and genitourinary toxicity persisted 60 months after RT and did not differ between the two dose schedules. Multivariate analyses revealed that the conventional schedule was of independent prognostic significance, not only for biochemical failure, but also for an increased risk of worse genitourinary symptoms at 4 years. A therapeutic advantage of the hypofractionated compared with the conventional dose schedule for RT of prostate cancer was evident at 90 months in the present study.