We conducted a population-based case-control study in King County, WA, to evaluate the association between usual caffeine intake and primary cardiac arrest. We identified primary cardiac arrest cases (N = 362) without a history of clinical heart disease or major comorbidity through paramedic incident reports during the period 1988-1994. We identified controls (N = 581), individually matched to cases on age (+-7 years) and gender and meeting the same general health criteria, through random digit dialing. We interviewed the spouses of cases and controls to obtain information on usual caffeine intake from coffee, tea, and cola during the prior year. After adjusting for cigarette smoking and other risk factors, we observed little association between daily consumption of the caffeine equivalent of fewer than 5 cups per day of drip coffee (=687 mg per day) was associated with a modestly elevated risk of primary cardiac arrest [odds ratio = 1.44; 95% confidence interval (CI) = 0.82-2.53]. The elevated risk associated with high caffeine consumption appeared to be restricted to never-smokers (odds ratio for >=687 mg per day = 3.2; 95% CI = 1.3-8.1).