Background: Whether adding an anti-androgen (AA) to a luteinizing hormone-releasing hormone (LHRH) agonist in the radiotherapeutic management of unfavorable-risk prostate cancer (PC) reduces all cause and PC-specific mortality (ACM and PCSM) risk is unknown. We evaluated whether the extent of AA received impacted the risk of ACM and PCSM within comorbidity subgroups. Methods: Between 1995 and 2001, 206 men with localized unfavorable-risk PC were enrolled on a randomized trial comparing radiation with (N = 102) or without (N = 104) 6 months of androgen deprivation therapy (ADT). Partial AA use (median: 4.2 months) occurred in 29 of the 102 men randomized to ADT due to discontinuation for elevated liver function tests or diarrhea. Cox and Fine and Gray's regressions were used to evaluate the impact of full versus partial AA use on PCSM and ACM risk within comorbidity subgroups adjusting for age and PC prognostic factors. Results: After a median follow-up of 16.62 years, 156 men died, 29 from PC (19%) including 25 (86%) and 4 (14%) PC deaths in the 157 and 49 men with no or minimal versus moderate to severe comorbidity, respectively. In men with moderate to severe comorbidity, an increased ACM risk with full as opposed to partial AA use that approached statistical significance was observed (adjusted hazard ratio (AHR), 2.25 [95% confidence interval (CI), 0.94-5.41]; P = 0.07); whereas only 1 and no PC deaths occurred in men receiving a partial versus a full AA course respectively. Among men with no or minimal comorbidity, no decrease in ACM (AHR, 0.97 [95% CI, 0.49-1.91]; P = 0.92) or PCSM risk (AHR 0.73 [95% CI 0.10-5.46]; P = 0.76) was observed for full versus partial AA use. Conclusions: Increasing AA use by 2 months does not appear to impact survival in men with localized unfavorable-risk PC and no or minimal comorbidity but may shorten survival in men with moderate to severe comorbidity raising concern regarding in whom and for how long the AA should be prescribed. Clinical trial information: NCT00116220