Background: The effect of dose-escalation with HDR-BT boost for high-risk PCa is not known. The objective is to compare 10-year PCa-specific mortality (PCSM) and overall mortality (OM) in non-metastatic patients treated with HDR-BT/EBRT (2004-2010) to EBRT alone (historical RCT, SPCG-7, 1996-2003). Methods: HDR-BT boosts (10 Gy x 2) were given 2 weeks apart followed by 50 Gy conformal EBRT (2 Gy x 25) to the prostate and seminal vesicles (assuming alpha/beta ratio of 3, EQD2 = 102 Gy). The HDR-BT/EBRT group (N:325) received Androgen Deprivation Therapy (ADT) for a total of 2 years. Patients in the control group (N:296) received 70 Gy (2Gy x 35) to the prostate and seminal vesicles with lifelong Anti-Androgen Treatment (AA). cT1-cT2 vs cT3 tumours and Gleason score 6-7 vs 8-10 were analysed. For each treatment group PCSM and OM were established by Kaplan-Meier (KM) analyses, and inter-treatment differences were tested by the logrank tests. Cox regression analysis evaluated the significance of available pre-treatment variables. Significance level p < 0.05. Results: In both groups the median age was 66 years. Median follow-up was 104 (range 13-120) and 120 (range 3-120) months for the HDR-BT/EBRT and EBRT groups respectively. KM plots revealed an 1.8% risk of PCSM in the HDR-BT/EBRT patient group and an 8.4% risk in the EBRT cohort (p = 0.001). For OM, the figures were 12.3% in the HDR-BT/EBRT group compared to 23.3% in the EBRT group (p = 0.014). In the Cox regression analysis, treatment (HR = 3.9, CI95% 1.8-8.3) and Gleason score (HR = 3.2, CI95% 1.8-5.9) were significantly associated with PCSM whilst T-stage, age and PSA levels were not. Treatment (HR = 1.7, CI95% = 1.1-2.6) was the only factor significantly associated with OM. Conclusions: In men with high-risk PCa dose-escalation with HDR-BT/EBRT compared to EBRT alone resulted in a significantly decreased risk of 10-year PCSM and OM despite shorter length of hormonal therapy. PCSM was significantly influenced by both Gleason score and type of treatment, whereas treatment remained the only significant covariate for OM.