The asthma mortality rate has increased steadily over the past 15 years in the United States and has only recently shown signs of leveling off. It is widely believed, although unproven, that many asthma deaths may be preventable. We have addressed one critical factor in severe asthma by attempting a definition of risk factors for intubation using demographic data and a retrospective cohort study of hospitalized asthmatic teenagers. This study included all asthmatics aged 13-19 years admitted over a 10-year period (1984-1994) to the University of California Davis Medical Center, Sacramento, California. A total of 143 such asthma admissions were reviewed, involving 68 females and 75 males, mean age 16.4 ± 2.3 years. Of this group, 85 teenagers were black, 34 were Caucasian, 14 were Hispanic, and 10 were Asian. By National Heart, Lung, and Blood Institute guidelines, there were 42 mild, 85 moderate, and 16 severe cases. Ten of the 143 teenagers studied required intubation for their asthma. The significant risk factors identified for intubation were active smoking and/or secondhand smoke exposure [odds ratio (O.R.) 21.3; 95% confidence interval (C.I.) 6.2, 73.1], crowding (O.R. 15.9; 95% C.I. 51.0, 50.1), atopy (O.R. 15.4; 95% C.I. 3.0, 78.3), psychosocial problems (O.R. 11.2; 95% C.I. 3.4, 36.9), language barrier (O.R. 7.4; 95% C.I. 2.1, 25.6), steroid dependence (O.R. 4.8; 95% C.I. 1.2, 18.7), low socioeconomic status (O.R. 5.5; 95% C.I.