OBJECTIVES: Inaccurate use of chronic medication is a serious problem for patients and health care professionals due to lower effectiveness of treatment. Non-persistence is likely to result in worsening of disease prognosis and increasing health care expenditures. The aim of the study was to investigate persistence with antihypertensive drugs. METHODS: A cohort study was conducted based on Dutch pharmacy dispensing records from IADB.nl. Enrolled were patients who initiated antihypertensive drugs in the period from 1st of January 1999 to 30th of June 2007. Patients with previous use of antihypertensive drugs 365 days prior to start, total supply <30 days, and patients aged <18 years were excluded. Antihypertensive drug classes included were diuretics, beta-blockers (BBs), calcium-channel blockers (CCBs), ACE-inhibitors (ACE), and angiotensin receptor blockers (ARBs). Persistence was measured by combining the proportion of days covered (PDC) of 80% and a gap period of 30 days (hybrid method). Both the PDC and gap period were varied in the sensitivity analysis. Cox proportional hazard analysis was conducted to calculate the unadjusted and adjusted (for age, gender, combi/mono therapy, and year of initiation) hazard ratios (HR) and 95% confidence intervals (95%CI) for the different antihypertensive drug classes. RESULTS: A total of 55,010 patients initiated antihypertensive drugs. Cox regression analysis showed statistically significant better persistence with ARBs as compared to other antihypertensives with adjusted HR of non-persistence of 1.33 (95%CI: 1.24-1.42), 1.76 (95%CI: 1.65-1.87), 1.98 (95%CI: 1.86-2.12), and 2.36 (95%CI: 2.19-2.53) for ACE, BBs, diuretics and CCBs, respectively. Variation in PDC and gap period had statistically significant influence on the survival, but did not influence the relative persistence of ARBs versus other antihypertensive drugs. CONCLUSIONS: Chronic users of antihypertensives are estimated to be most persistent with ARBs. Further results will be presented on the predictive value of current persistence on future adherence, as well as cost implications.