The impact of risk factors for acute myocardial infarction (AMI) strongly differs across populations and most studies do not consider age as an effect modifier. This study aims to estimate the population attributable fractions (PAFs) of established risk factors for non-fatal AMI, considering age stratification, within a population-based case-control study of Portuguese men. Cases were male patients consecutively admitted with an incident AMI, during 1999-2003 (n = 638) and controls were a representative sample of the non-institutionalized Porto, Portugal inhabitants (n = 851). PAFs were derived by the equation: PAF = 1 - Sigma (rho/R), in which rho is the proportion of cases in each exposure stratum and R is the adjusted odds ratio. PAFs were obtained for the individual effect of each factor and for combinations of them, after allowance for confounding. High waist-to-hip ratio (> 0.90), smoking and lower education levels (0-4 years) had the highest PAFs among men aged < or = 45 years: 81.2% (95% CI: 71.2-88.2), 63.5% (95% CI: 42.0-80.6) and 53.8% (95% CI: 40.9-66.2), respectively. For the oldest men, high waist-to-hip ratio (PAF = 88.7%, 95% CI: 77.6-94.7) and lack of leisure-time physical activity (PAF = 44.8%, 95% CI: 32.0-58.2) were the risk factors with the highest impact. Lifestyles explained 77.2% (95% CI: 53.4-90.9) of young myocardial infarction cases and 77.6% (95% CI: 65.3-86.4) of the cases aged > 45 years. Preventive targeted interventions to decrease the prevalence of such modifiable risk factors would likely reduce morbidity and mortality of cardiovascular events and related conditions.