Study objective: There is considerable uncertainty about the clinical features, respiratory physiology, and optimal management of patients with asthma requiring mechanical ventilation. Furthermore, the ventilatory and pharmacologic management of asthma requiring mechanical ventilation remain controversial. We hypothesized (1) that there are clinically identifiable and pathophysiologically different subgroups presenting with asthma requiring ventilation; (2) that lower dose steroid therapy (< 400 mg/d intravenous hydrocortisone) is adequate; (3) that permissive hypercapnia is safe; (4) that prolonged paralysis is generally unnecessary; and (5) that clinical outcome would be favorable in patients treated with this approach. Design: Review of medical records and intensive care charts and statistical analysis of findings. Setting: ICU of tertiary institution. Patients: Thirty-five consecutive cases of life-threatening asthma requiring mechanic ventilation. Results: Three clinical subgroups of ventilation-requiring asthmatics could be identified. Those presenting with steady deterioration (10), those with unstable asthma followed by a sudden 'dip' (16), and those with a sudden unexpected dip (9). Patients in the first group had a significantly lower PaCO2 (p<0.01) at presentation, but required ventilation for longer periods. Those in the second group had a significantly higher PaCO2 (p < 0.01) and required ventilation for a shorter period. Those in the third group had an intermediate PaCO2 level before intubation and the shortest period (p < 0.01) of mechanical ventilation. Five patients experienced their sudden dip after ingesting aspirin. Ten cases received 'high' dose hydrocortisone therapy (mean: 980 mg/24 h), and 25 received lower dose hydrocortisone (mean: 341 mg/24 h). No differences in illness severity at presentation or outcome could be detected between these two groups. Mean duration of ventilatory support was 36 h and mean duration of the ICU stay 52.1 h. Muscle relaxation was used in 12 patients for a mean period of 11.1 h. One patient was brain dead on arrival. All others survived. Conclusions: Life threatening asthma is an endpoint for several different clinical patterns of disease. No major clinical advantage could be found in our group of patients when high-dose steroids were used. Long-term use of muscle relaxants and prolonged mechanic ventilation are rarely needed in the management of patients with life- threatening asthma and excellent results can be achieved with a relatively simple management strategy.