The purpose of this study was to determine the total and cause-specific mortality in rheumatoid arthritis (RA) patients compared to a control population in northern Norway. One hundred and eighty-seven patients with RA and 930 population controls matched for age, gender and municipality were followed until death or for a maximum of 17 years. The total mortality in RA patients was twice that of their controls (MRR = 2.0, 95% CI = 1.6-2.5). Patients possessing serum rheumatoid factors did not have a higher relative mortality than the seronegative patients. There was no statistically significant increased mortality from cancer or cardiovascular diseases. Indications for a higher death rate in RA patients than in controls were found for infection and sudden death.