Pain following intravenous injection as well as thrombophlebitis are substantial side effects of etomidate that have been reported from the first clinical study (1972-1973) onwards. Investigations of our own and by Gran et al. have pointed out that injectable etomidate with intralipid as a solvent removes side effects without impairing the good hypnotic action. The idea of using a lipid emulsion as a solvent was presented a few years later, inducing two further studies. METHOD. Both pharmacodynamic (continuous EEG registration) and pharmacokinetic [determination (HPLC) of plasma levels of the active substance] investigations were carried out on volunteers. At random 16 volunteers received etomidate in propylene glycol or etomidate in lipid emulsion for general anesthesia. A dose of 0.3 mg kg-1 was given over 60 s. In a clinical study 100 patients were divided into two independent groups of 50 each. They received either the commercially available etomidate in 35% propylene glycol (group I) or the new formulation containing 20 mg etomidate in 10 ml of a lipid-emulsion (Lipofundin MCT 20%) (group II). A dose of 0.3 mg kg-1 etomidate was given. RESULTS. There was a higher concentration of etomidate for 8 min after injection in lipid emulsion compared with etomidate in propylene glycol. The plasma concentration of etomidate (200 ng/ml) correlates with C2 corresponding to the light sleep stage after etomidate in propylene glycol, but with D0 according to the deep sleep stage after etomidate in lipid emulsion. At lower plasma concentrations, the hypnotic action of etomidate in propylene glycol is stronger than the effect of the lipid emulsion. This result means that it is possible that a part of the etomidate remains in the lipid particles. In the clinical study the anesthetic induction time was nearly identical in both groups; blood pressure and heart rate were stable. Following etomidate in propylene glycol, 36% of the patients complained of a painful injection. On the first postoperative day, 9 of 47 patients showed signs of phlebitis and three others thrombosis. On the 7th day a venous reaction was evident in 22% of these patients; 2 patients had developed phlebitis, 5 thrombosis and 4 thrombophlebitis. After etomidate in lipid emulsion, there were no signs of local irritation. The same results have been obtained in the study with volunteers. CONCLUSION. Two unpleasant side effects of etomidate, pain on injection and postoperative thrombophlebitis, were abolished by the solvent "lipid emulsion".