The prognosis of patients with COPD requiring admission to the ICU is generally believed to be poor. There is a paucity of long-term survival data. We undertook a study to examine both the in-hospital and 5-year mortality rates and to identify the clinical predictors of these outcomes. We conducted a retrospective cohort study of 57 patients admitted to the ICU between January 1999 and December 2000 for acute respiratory failure attributable to COPD. The mean (+/-SD) age of the study population was 70 +/- 8 years. More than 90% of patients required intubation, and the mean duration of mechanical ventilation (MV) was 2.3 +/- 2.2 days. The in-hospital mortality rate for the entire cohort was 24.5%. The mortality rates at 6 months and 1, 3, and 5 years were 39.0%, 42.7%, 61.2%, and 75.9%, respectively, following admission to the ICU. The median survival time for all patients was 26 months. The mortality rate at 5 years was 69.6% for patients who were discharged alive from the hospital. Using multivariate analysis, hospital mortality correlated positively with age, previous history of MV, long-term use of oral corticosteroids, ICU admission albumin level, APACHE (acute physiology and chronic health evaluation) II score, and duration of hospitalization. No factors predictive of mortality at 5 years were identified. We support previous findings of good early survival and significant but acceptable long-term mortality rates in patients who have been admitted to the ICU for acute exacerbation of COPD. Increased age, previous history of MV, poor nutritional status, and higher APACHE II score on ICU admission could be identified as risk factors associated with increased mortality rates. Long-term survival of patients with COPD who required MV for an acute exacerbation of their disease cannot be predicted simply from data available at the time of intubation. Physicians should incorporate these factors in their decision-making process.