BACKGROUND: To examine trends in 3-year mortality after a first hospitalization with diagnosed atrial fibrillation in a large cohort with and without important comorbidities. METHODS: The Swedish Hospital Discharge and Cause of Death Registries were linked to investigate trends in mortality for all patients 35 to 84 years hospitalized for the first time with a discharge diagnosis (principal or contributory) of atrial fibrillation in Sweden during 1987 to 2006.We performed an analysis of temporal trends in mortality stratified for presence or absence of co-morbidities affecting survival. RESULTS: Exactly 376,000 patients (56% male, mean age 72 years) with a first diagnosis of atrial fibrillation during 1987–2006 were identified and followed for 3 years. Patients with one or more of the prespecified comorbidities had the highest mortality and the largest absolute decline in mortality, but patients without these comorbidities had a slightly larger relative decline (absolute risk reduction in 3-year mortality (AAR) from 42.5 to 34.7%, Hazard Ratio (HR) 0.76; 95% confidence interval (95% CI) 0.74 to 0.77 versus ARR 16.2% to 11.7%, HR 0.71; 0.68 to 0.74. In patients aged below 65 years,with no comorbidities, there was minimal change inmortality, and they still had a 2 times increased mortality compared to the general population (SMR 1.95; 1.84-2.06). CONCLUSIONS: Survival after a first hospitalization with a diagnosis of atrial fibrillation improved regardless comorbidities. Patients aged < 65 years old without diagnosed comorbidities still had a poor prognosis compared to the general population.