Fifteen patients with chronic renal failure (CRF) were given midazolam 0.2 mg/kg iv over 15 s. All but one lost consciousness in a time ranging from 22–100 s (mean ± SD was 55 ± 26 s) after drug administration. Patients regained consciousness from 6–105 min (mean 53 ± 32) after drug administration. The calculated mean plasma level of midazolam at arousal was 81 ± 47 ng/ml. Pharmacokinetics parameters were determined from midazolam plasma levels measured in 16 consecutive venous blood samples. The pharmacokinetic parameters in CRF patients were compared with those of healthy volunteers matched for age, sex, and body size with the CRF patients. Protein binding was determined by equilibrium dialysis. CRF patients had a significantly higher (P ≤ 0.005) plasma-free drug fraction (6.5% ± 0.7) compared with the control patients (3.9% ± 0.1). Total (bound plus unbound) kinetics differed in the two groups: volume of distribution 3.8 ± .3 1/kg in CRF patients versus 2.2 ± .2 1/kg in controls (P ≤ 0.001), and clearance 11.4 ± 1.6 ml·min-1-1·kg-1 in CRF patients versus 6.7 ± 0.9 ml·min-1kg-1 in controls (P ≤ 0.02). When kinetic parameters were corrected for protein binding, CRF patients unbound volume of distribution (63.5 ± 6.8 1/kg) and free drug clearance (189 ± 29 ml·min-1·kg-1) were not different from the control group's volume of distribution (55.6 ± 5.7 1/kg) and free drug clearance (176 ± 24 ml·min-1·kg-1). Midazolam elimination half-life was almost identical in both groups; in CRF it was 4.58 ± 0.75 h and 4.93 ± 1.08 h in healthy controls. Because CRF does not alter the distribution, elimination, or clearance of unbound midazolam, changes in the pharmacodynamic profile of midazolam in CRF patients, if they exist, are more likely due to inherent alterations in drug sensitivity than to pharmacokinetic changes.