In the Candesartan Antihypertensive Survival Evaluation in Japan (CASE-J) trial, comparable efficacy was noted between candesartan and amlodipine in the incidence of cardiovascular (CV) morbidity and mortality during 3.2 years of follow up. Candesartan suppressed new-onset diabetes more effectively than amlodipine. In this observational study, we investigated whether or not the efficacy of the two drugs is sustainable for another 3 years beyond the experimental period of the CASE-J trial. Of the 4728 high-risk hypertensive patients initially enrolled in the CASE-J trial, 2232 agreed to further follow up. The primary endpoint was a composite of CV morbidity and mortality. The distribution of demographic characteristics for the 2232 patients in the CASE-J extension was similar to that in the initial 4703 patients in the CASE-J trial. Both drugs controlled blood pressure well over the relatively long period of time. The incidence of CV events was 15.5/1000 patient years in the candesartan group and 16.3/1000 patient years in the amlodipine group (Hazard ratio (HR)=0.95, 95% confidence interval (CI)=0.77-1.18; P=0.650). The incidence of new-onset diabetes was significantly lower in the candesartan group (9.5/1000 patient years) than in the amlodipine group (13.3/1000 patient years), representing a 29% risk reduction for new-onset diabetes (HR=0.71, 95% CI=0.51-1.00, P=0.0495). In conclusion, candesartan and amlodipine showed comparable efficacy against CV events beyond the experimental period of the CASE-J trial in high-risk hypertensive patients. In addition, the effects of candesartan on new-onset diabetes observed during the CASE-J trial were sustained in the CASE-J extension. The CASE-J extension, which covered a 3-year extension of follow-up from the original trial, corroborated the results of the CASE-J trial.