A case-control study has previously been reported of asthma deaths in people aged 5-45 years who had a hospital admission for asthma (the index admission) in New Zealand during 1981-1987. The study has been re-analysed to examine the association between markers of asthma severity and risk of asthma death or hospital admission; patients prescribed fenoterol were excluded from this re-analysis because of the previously reported interaction between fenoterol, asthma severity, and asthma deaths. The re-analysis included 39 patients who died of asthma during the 12 months after their index admission, 226 patients who had a readmission for asthma during the 12 months after their index admission, and 263 controls chosen from all index admissions. An admission in the previous 12 months was the strongest marker of subsequent risk of death (odds ratio (OR) = 3.5, 95% confidence interval (CI): 1.8-6.9, P less than 0.01), and was also a strong marker of subsequent risk of readmission (OR = 3.0, 95% CI: 2.1-4.2, P less than 0.01); the risk increased with the number of previous admissions. Three or more categories of prescribed asthma drugs was also associated with subsequent death (OR = 1.7, 95% CI: 0.9-3.3, P = 0.13) or readmission (OR = 1.9, 95% CI: 1.3-2.7, P less than 0.01); prescribed oral corticosteroids was only weakly associated with subsequent death (OR = 1.3, 95% CI: 0.6-2.8, P = 0.59), but was more strongly associated with subsequent readmission (OR = 1.9, 95% CI: 1.2-2.8, P less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)