Background Atrial fibrillation ( AF ) prevalence is increasing, and body mass index ( BMI ) is a risk factor for AF . However, sex differences in the impact of BMI on AF risk have not been fully elucidated. Methods and Results Data from the fourth survey (1994–1995) of the Tromsø Study (Norway) were used to investigate the association of single‐measurement BMI on future AF risk. To analyze the influence of BMI changes on AF risk, data from individuals who attended the third and fourth study surveys were used. AF diagnosis was derived from record linkage and end point adjudication. Cox regression analysis was conducted using fractional polynomials of BMI and BMI change with models adjusted for age, baseline BMI (change analyses), risk factors, comorbidities, and antihypertensive medications. Data were available for 24 799 individuals from the fourth survey (mean age, 45.5±14.2 years; 52.9% women). Over 15.7±5.5 years, 811 women (6.2%) and 918 men (7.9%) developed AF . In men, lower BMI decreased AF risk and higher BMI increased risk (hazard ratios [95% confidence intervals] for BMI 18 or 40 kg/m ² compared with 23 kg/m ² were 0.75 [0.70–0.81] and 4.42 [3.00–6.53], respectively). The same pattern was identified in women. Two surveys were attended by 14 652 individuals. In men and women, a decrease in BMI over time was associated with decreased AF risk and an increase in BMI was associated with increased AF risk. Conclusions Within a population cohort, BMI was positively associated with AF risk. Change in BMI over time influenced AF risk in both men and women.