BACKGROUND Lisinopril and perindopril are two commonly used first-line antihypertensive agents. Few studies compared their effectiveness in reducing the incidence of renal diseases and diabetes. METHODS Adult patients who received new prescriptions of lisinopril or perindopril from 2001 to 2005 in all public hospitals and clinics in Hong Kong were included, and followed up for at least 2 years. Patients prescribed the angiotensin converting enzyme inhibitors (ACEIs) for <1 month were excluded. The incidence of admissions due to renal diseases and diabetes was evaluated. We used Cox proportional hazard regression models to assess hospital admissions as the outcome measures, adjusting for age, sex, socioeconomic status, service types, and the proportion of days covered as a measure of medication adherence. The regression models were constructed with propensity score matching to minimize indication biases. RESULTS 20,252 eligible patients with an average age of 64.5 years (SD 15.0) were included. The admission rates 24 months within the date of index prescription due to renal diseases were 3.1% (lisinopril) and 2.3% (perindopril); and 9.6% (lisinopril) and 7.2% (perindopril) for diabetes. Except for admissions due to diabetes at 6 months, lisinopril users were significantly more likely to be admitted due to renal diseases (adjusted hazard ratios: 1.304 to 1.378) and diabetes (1.146 to 1.231) than perindopril users at all time points. CONCLUSIONS Patients prescribed different ACEIs might have a different incidence of hospital admissions. Future studies should be conducted to evaluate the comparative effectiveness of different ACEIs on various patient-centered outcomes by head-to-head randomized controlled trials.