BACKGROUND: The effects of aspirin in subjects without cardiovascular disease are controversial. In the intensively treated patients of the Hypertension Optimal Treatment (HOT) Study, randomization to low-dose aspirin (75 mg daily) versus placebo significantly reduced cardiovascular events (-15%) and myocardial infarction (-36%), but increased major bleedings (+65%). The present analyses of HOT Study data aim at identifying subgroups of hypertensives with different benefit-to-harm ratios from aspirin, in order to provide recommendations about the use of aspirin in hypertension. METHODS: The 18 790 hypertensive patients (aspirin 9399, placebo 9391; average treatment duration 3.8 years) were stratified for global cardiovascular risk and for individual risk factors. Subgroup-treatment interaction analyses (end points: cardiovascular events, myocardial infarction, major bleedings) were performed by a Cox proportional hazard model. Relative and absolute benefits and harms were calculated. RESULTS: Interaction analyses indicated that of all subgroups, only patients with serum creatinine > 1.3 mg/dl had a significantly greater reduction of cardiovascular events and myocardial infarction (-13 and -7/1000 patient-years), while risk of bleeding was not significantly different between subgroups. In addition to patients with higher creatinine, a favourable balance between benefit and harm of aspirin was found in subgroups of patients at higher global baseline risk and baseline systolic pressure > or = 180 or diastolic pressure > or = 107 mmHg. CONCLUSIONS: Low-dose aspirin should be recommended to well-treated hypertensive patients with even moderate increase in serum creatinine. Aspirin may also be recommended in well-treated hypertensives at higher global cardiovascular risk or higher initial blood pressures.