Patients with rheumatoid arthritis (RA) have a substantially reduced life expectancy. The standardized mortality ratio in different studies has ranged from 1.13 to 2.98. This mainly applies to rheumatoid factor (RF)-positive cases, although there is a subgroup of RF-negative cases with an adverse long-term prognosis. Clinically based studies probably overestimate the true shortening of life span and population-based studies may underestimate it. Excess mortality from infection and from renal disease likely reflects the presence of severe disease, whereas most of the added mortality from gastrointestinal causes is treatment related. The reasons for the surplus of mortality from cardiovascular causes are not fully known. RF may have a direct role, and preillness factors such as smoking may predipose patients to RA and also render them susceptible to cardiovascular diseases. The excess mortality associated with RA is appreciably higher than is apparent from the cases in which RA is regarded as an underlying cause of death. The effect of treatment on mortality remains largely unknown.