Midazolam, a new benzodiazepine was compared with diazepam as intraoperative sedative during spinal anaesthesia in a double blind study. Eighty five patients participated in this clinical trial (placebo 25, midazolam 30, diazepam 30). Drug treatment was administered after spinal anaesthesia at 5 min of supine position. Level of spinal anaesthesia in terms of sympathetic blocade was similar in aU the patients. Dosage of drug were tailored to the needs of each patient till the end point of undisturbed sleep. Intraoperative sedation based on predetermined criteria, recovery by Trieger's test at 4 hr after the drug administration, antegrade amnesia by memory card and haemodynamics were assessed. Mean doses of 240 Ilg kg-! and 500 Ilg kg-Iof midazolam and diazepam were required to induce sleep which suggest an equivalent dose ratio of midazolam: diazepam as 1: 2.1. Both the drugs produced effective sedation during spinal anaesthesia, 86.6% patient with midazolam and 93.3% patients with diazepam had deep undisturbed sleep. Clinical recovery was fast in midazolam administered patients but did not meet the criteria of early discharge by Trieger's test at 4 hr after the administration of the drug. There were no gross changes in the circulatory effects of the two drugs. Decrease in mean systolic pressure (<10%) and marginal increase in heart rate was seen. The frequency of antegrade amnesia was greater with midazolam. Global assessment as judged by anaesthesiologist and patients satisfaction for anaesthesia and recovery was higher (93.3%) with midazolam as compared to (79.9%) diazepam. Trieger test consists of a series of dots arranged in a simple pattern which the patient connects as accurately as possible. The accuracy and timings were compared with the baseline test carried out before premedication. Decrease in accuracy was measured by recording the number of dots missed and the cumulative distance of miss in mm, compared with the baseline test. Position of the patient, moderator staff, place, material and light were the same during performance of the test.Patients were premedicated with injection atropine 0.6 mg and injection fortwin 0.5 mg kg -I intramuscularly one hour prior to surgery in the anaesthetic room and iv cannula was placed on the dorsum of the hand; heart rate, arterial pressure, level of sedation were recorded.Lumbar puncture was performed in the sitting position at the 3rd interspace and 0.5% plain bupivacaine 4 ml was injected at 1 mi. 5 Sec -I. The patients were returned to the supine position, heart rate, blood pressure and upper level of block (pinprick test of analgesia) were recorded every 5 min until spread ceased. Respiratory rate was also recorded. Drug (placebo, midazolam and diaze-