Objective To explore and appraise the representation of samples in clinical trials of hypertension medications. Methods Four databases (MEDLINE, CBM, CNKI and VIP )were searched systematically. All RCTs of hypertension medication run in China from January 1996 to December 2006 were collected, and then appraised by a quality assessment checklist. Those RCTs with 3 or more grading score were enrolled for critical appraisal of representation of samples. Results A total of 164 RCTs on hypertension were targeted and reviewed systematically. 122 (74.4%) RCTs selected secondary or tertiary cares as their trial settings. 81 (49.4%) RCTs described the source of samples, in which 86.3% were hospitalized patients. 16.7% patients in 12 (7.3%) RCTs were excluded in run-in periods, 24.6% patients were ineligible for inclusion criteria in 8 (4.9%) RCTs. 8.4% eligible patients were non-enrolled in 3 RCTs. 89.4% subjects were mild to moderate essential hypertension. The proportions of women and the patients aged 65 years or older were 41.1% and 16.9%, respectively, and were markedly lower than that in community (Plt;0.001). Conclusion Representation of samples in clinical trials of hypertension medications was not good and the influencing factors of representation were not reported sufficiently. These would undermine the efforts to provide evidence-based care to all hypertensive patients.