The effects of continuous and supplementary bolus dose administration of etomidate have been investigated in ten artificially ventilated patients in traumatic coma. Continuous infusion of etomidate (5-25 micrograms/kg/min) proved to be a practical and safe means of sedating these patients and appeared to control moderately elevated ICP. Additional bolus doses of etomidate (0.2 mg/kg) always reduced acutely elevated ICP (greater than 20 mmHg), which fell by a mean of 33%. However, MAP usually fell, and occasionally serious hypotension occurred. Of a total of 61 bolus dose administrations which were analysed, CPP rose on 40 occasions, fell on 19 and was unchanged twice. There was a weak correlation between the control level of ICP and the magnitude of the fall in ICP in response to the bolus dose of etomidate (r = 0.51, p less than 0.001). Bolus doses of etomidate given just before noxious stimulation, for example chest physiotherapy, prevented or limited the expected rise in ICP (with bolus mean change in ICP = -2.7 +/- 6.9 mmHg, without bolus mean change in ICP = +7.0 +/- 6.4 mmHg). Again MAP tended to fall following the bolus dose. Overall CPP tended to fall slightly following stimulation whether or not a bolus dose was administered (-3.2 +/- 11.1 mmHg and -4.9 +/- 11.5 mmHg respectively). However, when the bolus of etomidate was not given, occasional dramatic and dangerous rises in ICP were seen, in spite of the infusion, during which CPP fell to critical levels. This very rarely occurred when the bolus had been given. Furthermore, serious episodes of hypotension in response to etomidate administration appeared to occur mainly in patients who were relatively hypo-volaemic.