Premedication with diazepam, hydroxyzine or secobarbital significantly increased ketamine-induced sleep time (137 +/- 3.5 min, 135 +/- 9.2 min, 128 +/- 4.7 min) over that of unpremedicated controls (98.5 +/- 4.4 min) in man. The corresponding mean plasma half-lives (t1/2) of ketamine were longer in patients premedicated with diazepam or seconbarbital (57.8 +/- 4.9 min, 46 +/- 3.2 min) than in controls (36 +/- 1.8 min). Ketamine t1/2 in the perfusate of isolated, perfused rat livers was prolonged 30 to 50 per cent by addition of diazepam, secobarbital, or hydroxyzine. The data suggest that these commonly used premedicants decrease the rate of ketamine metabolism.