To explore the relationship between vigorous physical activity and primary cardiac arrest (PCA), we identified, through emergency service incident reports, 163 cases of PCA aged 25 to 75 years. Control subjects matched for age, sex, residence, and the absence of prior clinical heart disease or major comorbidity were identified from the community. Spouses of subjects were interviewed to quantify leisure-time activity (LTA) during the prior year. Energy expended in high-intensity LTA, requiring 60% of maximum oxygen intake, was determined. The risk of PCA was 55% to 65% lower in persons in the two upper quartiles of high-intensity LTA than in persons without high-intensity LTA (95% confidence interval of relative risk, 0.22 to 0.85). Because this association was demonstrated in a clinically healthy population without prior morbidity, our data support the hypothesis that high-intensity LTA protects against PCA.