To evaluate the effect of end-stage renal insufficiency and haemodialysis on the elimination of lorazepam, single oral doses of the drug (2.5 mg) were administered to normal subjects and patients with chronic renal failure (CC(r) : less than 2 ml/min) in the interdialysis period and during haemodialysis. The concentration of lorazepam and its major metabolite, lorazepam-glucuronide, were assayed using electron capture g.l.c. Plasma half-life (T1/2) of unchanged lorazepam in the patient group (11.3 +/- 0.6 h) was not different from that obtained in normals (11.1 +/- 0.9 h). Only minor quantities of the unchanged drug could be recovered in the 24 h urine in both groups: 0.3% of the ingested dose in normals and trace amounts in the patient group. No unchanged lorazepam could be detected in the ultrafiltrate from the coil kidney. Since the lower sensitivity of the method is about 5 ng/ml, this would indicate the in vivo binding of the active drug to plasma proteins to be at least 70%. The effect of haemodialysis on lorazepam plasma T1/2 was also insignificant (9.4 +/- 1.0 h). Urinary excretion of lorazepam-glucuronide was found to be considerably decreased in chronic renal failure associated with accumulation of high concentrations of this conjugate in plasma during days after a single oral dose. The plasma T1/2 of this conjugate in normals was 20.7 +/- 2.1 h. Roughly 35% of this main metabolite's concentration in plasma was detected in the ultrafiltrate from the coil kidney indicating the dialyzability of this conjugate and that the extent of plasma protein binding of lorazepam-glucuronide in vivo was approximately 65%. The above results indicate that after a single oral dose (2.5 mg) the biotransformation of lorazepam to its glucuronide conjugate remains unaltered and that high concentrations of this metabolite accumulate in plasma in the presence of severe renal function impairment.