To evaluate risk factors for asthma mortality, an unmatched case-control study was undertaken in the Canadian prairie provinces of Alberta, Saskatchewan, and Manitoba. Those between the ages of 5 and 50 (inclusive) who died from an acute exacerbation of asthma were compared to a control group of people with asthma from the same geographical areas who were contacted using random-digit dialing. Because no deaths occurred among residents less than 15 years old, this analysis was limited to cases and controls between 15 and 50 years old. Of the 38 deaths that occurred between November 1992 and October 1995, data were obtained from next of kin for 35 (92.1%). Of the 210 potential controls that were identified, 142 returned completed questionnaires (67.6%). Cases were more likely than controls to have asthma reported to be severe, to have experienced nocturnal symptoms, to have had cardiopulmonary resuscitation (CPR)/intubation, and to have had more healthcare utilization in the previous year. Medication use was also more common among cases compared to controls. Specific asthma triggers were reported more often for cases than controls; weather changes, excitement, depression, and stress showed the greatest case control differences. Although a number of very strong risk factors for death from asthma were identified, death from asthma is so rare in this age group that it is not possible to label an individual as "likely" to die from asthma. Nonetheless, patients, caregivers, and health professionals should be aware of indicators that would suggest greater risk.