The pharmacokinetics of propofol, 2,6 diisopropylphenol, were compared in 6 end-stage renal failure patients aged 66.3 +/- 12.1 years and in 5 normal patients aged 45.5 +/- 13.5 years. Anesthesia was induced with propofol (2 and 2.5 mg.kg-1 respectively) and fentanyl (0.1 mg). Anesthesia was maintained with propofol (9 and 10 mg.kg-1 x h-1 respectively). Patients breathed spontaneously a 50 per cent oxygen in nitrous oxide mixture. Two ml blood samples were taken during anesthesia and at regular intervals until up to 24 hours after infusion. Plasma levels of propofol were determined by HPLC with electrochemical detection. Propofol conjugates were determined after hydrolysis with beta glucuronidase or HCl. Results were expressed as median [lower-upper percentiles]. Propofol clearance (1.53 [1.02-2.10] L.min-1 x kg-1 versus 1.65 [1.39-1.78] L.min-1 x kg-1 in normal patients), and half-lives elimination were not modified by renal failure. Renal failure patients exhibited a higher volume of distribution at steady state as compared to normal patients (19.28 [11.71-76.81] L.kg-1 versus 8.60 [6.58-9.81] L.kg-1, p < 0.05). Renal failure did not affect the production of conjugates but they accumulated in blood of renal failure patients. Despite this, no difference in time to eyes opening and propofol concentration were observed, confirming the absence of clinical effect of these metabolites.