This prospective randomized study compared unilateral and bilateral spinal anesthesia with respect to intraoperative and postoperative complications, and time to discharge from hospital for knee arthroscopies in outpatients. We studied 70 ASA I patients scheduled for elective outpatient knee arthroscopy. The patients were randomly allocated into two groups to receive either 3 ml (15 mg) 0.5% hyperbaric bupivacaine (bilateral group) or 1.5 ml (7.5 mg) 0.5% hyperbaric bupivacaine (unilateral group). The duration of motor and sensory block and the time to discharge from the hospital were all recorded. Perioperative complications such as hypotension, bradycardia, nausea, vomiting, urinary retention, if present, were recorded. The patients were interviewed by telephone 7 days later, and each patient was asked about headache or backache. The duration of motor and sensory block, and the time to discharge from hospital was shorter in the unilateral group than in the bilateral group. Three patients in the bilateral group were treated for hypotension. Bradycardia occurred in two patients in the bilateral group, and three patients required temporary bladder catheterization due to delay in recovery of spontaneous urination. Nausea and vomiting occurred in three patients in bilateral group. Nine patients in the bilateral group and six patients in the unilateral group developed postspinal headache. Backache occurred in five patients in the bilateral group and in six patients in the unilateral group. Our data indicate that the use of unilateral spinal block is a suitable technique for knee arthroscopies in outpatients.