We studied the profiles of all-cause and cardiovascular (CVS) mortality among users of different antihypertensive classes in a Chinese population. From electronic patient records, a cohort study was conducted among 18,338 patients who ever newly prescribed an alpha-blocker, thiazide diuretic, beta-blocker, calcium channel blocker (CCB) or agents acting on the renin-angiotensin system (RAS) without drug discontinuation or switching in the public primary-care sector in a large Territory of Hong Kong during January 2004-June 2007. The odds ratios of mortality (all-cause and CVS) were evaluated according to the prescribed antihypertensive drug classes by Cox proportional hazards regression analyses. A total of 823 deaths (4.5 were reported during the study period. The crude proportions of all-cause mortality were highest in alpha-blockers (6.2 and CCB (5.7, but lowest in beta-blockers (2.8. Compared with CCB, patients on thiazide diuretics were shown to have statistically significantly lower all-cause (adjusted hazard ratios (aHRs) 0.75, 95% CI 0.60, 0.93, P=0.010) and CVS mortality (aHR 0.40, 95% CI 0.21, 0.78, P=0.007), but the 95% CI of the odds ratios of the major drug classes overlapped. When each drug class was used as a reference group, or when patients with only uncomplicated hypertension were included, their respective 95% CI similarly overlapped. Antihypertensive drug classes were associated with statistically comparable odds of all-cause and CVS mortality. This finding from real-life clinical practice further supports the position statements from international guidelines, which recommend that the major antihypertensive drug classes are suitable for initiating pharmacotherapy for the management of hypertension.