We examined the prevalence of atrial fibrillation (AF) in a large French population according to age, risk factors, all-cause mortality, and cardiovascular and cerebrovascular mortality. The study population was composed of 98,961 men and 55,109 women over 30 years of age who had a free medical checkup at the IPC Center (Centre d'Investigations Préventives et Cliniques). Routine electrocardiograms revealed the presence of AF in 235 men (mean age 60.2 +/- 10.3 years) and 63 women (mean age 62.5 +/- 9.1 years). Mean follow-up was 15.2 years. The relative risk of death [Hazard Ratio (95% CI)] was determined with a Cox regression model. The prevalence of AF increased strongly with age in both genders and was higher among men. Before 50 years of age, AF was present in 0.05% of men and 0.01% of women, compared to 6.5% and 5.2%, respectively, in over-80s. After adjustment for age, factors significantly associated with AF were cardiopathy [Odds Ratio (OR) = 3.2 (2.3-4.5) among men and 4.9 (2.5-9.5) among women], hypertension [OR = 1.4 (1.1-1.9) in men and 2.2 (1.2-3.9) in women], overweight [OR = 2.2 (1.4-3.2) in men and 2.3 (1.0-5. 1) in women], ventilatory failure [OR = 1.4 (0.9-2.2) in men and 4.9 (2.4-10) in women], diabetes [OR = 1.7 (1.1-2.5) in men] and alcohol consumption [OR = 1.7 (1.2-2.4) in men]. The relative risk of death was then adjusted for age, cardiopathy, left venticular hypertrophy, blood pressure, cholesterol, glycemia, body mass index, smoking, alcohol, and vital capacity. The HR of all-cause mortality was 1.5 (1.0-2.0) in men and 1.8 (1.0-3.3) in women. The HR of cardiovascular mortality was 2.2 (1.2-3.1) in men and 3.4 (1.5-7. 7) in women, while for stroke-related mortality it was 2.0 (0.7-4.3) in men and 4.5 (1.3-16) in women. No association was found between AF and non-cardiovascular mortality in either men or women. The risk of death among men without cardiopathy or hypertension, after adjustment for the other risk factors, was not significantly increased (overall mortality 1.1 (0.5-2.0), cardiovascular mortality 1.4 (0.6-2.9)). In contrast, men with cardiopathy or hypertension had an adjusted HR of 1.7 (1.1-2.8) for overall mortality and 2.6 (1.3-5.3) for cardiovascular mortality. In conclusion, after adjustment for all risk factors, the AF-related relative risk of overall mortality and of cardiovascular mortality was higher among women than among men, especially for cerebrovascular mortality. AF was not an independent risk factor for death among men free of cardiopathy and hypertension.