OBJECTIVE: Adherence to antihypertensive drug treatment is suboptimal. The present study investigates the effect of early treatment discontinuation with antihypertensive drugs on the risk of acute myocardial infarction (AMI) or stroke in daily clincal practice. METHODS: In the PHARMO Record Linkage System, which includes all records of drug dispensings and hospitalisations for > or = 2 million subjects in the Netherlands, new users of antihypertensive (AHT) drugs > or = 18 years of age were studied during the period 1 January 1993 – 1 October 2002 to determine the risk of AMI or stroke related to persistence with AHT. Patients were initially followed for 2 years to determine persistence with AHT, and then for a further 2 years or until the firsthospital admission for AMI or stroke, death, or end of the study period. Patients using AHT for secondary prevention of cardiovascular disease were excluded from the study cohort. RESULTS: The study cohort included 77 193 AHT users. The percentage of non-persistent patients was 55% at 2 years, with the lowest non-persistence rates for angiotensin-receptor blockers (ARBs) and ACE-inhibitors (40%) and the highest rates for beta-blockers, calcium-channel blockers (CCBs) and diuretics (54-61%). Non-persistent AHT use was associated with a 15% increased risk of AMI (RR 1.15; 95% CI 1.00-1.33) and a 28% increased risk of stroke (RR 1.28; 95% CI 1.15-1.45). CONCLUSIONS: The results of this study show that in daily clinical practice early discontinuation of antihypertensive drug treatment in primary prevention increases the risk of subsequent AMI or stroke.