The rate of death from asthma in the United States has been increasing over the last decade, but such deaths still remain uncommon. Mortality rates differ markedly by region and ethnicity, and case-fatality rates are highest in black men living in inner cities. In general, patients in greatest jeopardy are those with severe, unstable disease who are not being objectively monitored; however, death can occur in anyone if the attack is intense enough. Sudden catastrophic episodes of asthma occur but are very rare. Failure to recognize the seriousness of the terminal episode or to treat the episode appropriately remains the chief contributing cause of poor outcome. Little compelling evidence shows that adverse effects of medications play much of a role in asthma-related death. The disease characteristics that place patients at risk remain inadequately defined. Of those suggested, only a history of recurrent hospitalization and the need for ventilatory assistance are specific enough to be helpful. These characteristics, however, are found in only 36% and 6% of cases, respectively; thus, their absence is of no value in assessing risk. The prognosis after a near-fatal episode of asthma is poor, and approximately 10% of patients die in the year after the event. Given the current state of knowledge, it is wise to view all exacerbations of asthma that last longer than a few days as potentially fatal and to treat them accordingly. This is especially true in patients who have previously had a severe episode of asthma.