The objective of the present study was to investigate the long-term prognosis of near-fatal asthma. A retrospective cohort study design was used. Cases were defined as any asthmatic individual requiring mechanical ventilation for the first time for an asthma exacerbation between January 1, 1983 and December 31, 1988. The consecutive sample of patients was drawn from four study sites, specifically four intensive care units (ICU), based in a large urban area (1 million inhabitants). These four ICU total approximately 5,000 admissions per year and are the referral centers for more than 95% of patients requiring respiratory intensive care in the area. Data collection was obtained by questionnaires addressed to the patients and to their attending physicians and was completed by telephone calls if necessary. A total of 147 patients entered the study. The long-term outcome could be evaluated in all but two patients. The follow-up period ranged from 1 to 75 months. In-hospital mortality was 16.5%. Among the 121 patients discharged from the ICU, 18 subsequently died, 17 of whom died from a new attack of asthma. Post-hospitalization mortality was 10.1% (95% CI, 5.9 to 16.8%) after 1 yr, 14.4% (CI, 9 to 22.3%) after 3 yr, and 22.6% (CI, 12.7 to 36.8%) after 6 yr. Nearly two-thirds (61.5%) of these secondary deaths occurred within the year following discharge from the ICU. Smoking was associated with a higher in-hospital mortality, as well as with a higher posthospitalization mortality. Age was also independently associated with a higher posthospitalization mortality. It is noteworthy that the secondary deaths were mostly observed in patients over 40 yr of age.(ABSTRACT TRUNCATED AT 250 WORDS)