We compared general anesthesia (GA) with spinal anesthesia (SA) for outpatient knee arthroscopy. Fifty patients were randomized to receive either sevoflurane in a mixture of nitrous oxide (60%) in oxygen with laryngeal mask, or 30 mg of lidocaine 1% spinal anesthesia. All patients received premedication with intravenous (iv) lornoxicam 8 mg and at the end of the operation; patient's knee joint was injected with 1 ug/kg clonidine diluted in 20 ml of 0.25% bupivacaine. Postoperatively, iv fentanyl was given if visual pain scale (VAS) at rest exceed 4 and on discharge from the hospital, patients were instructed to take lornoxicam 8 mg tablet every 12 hour (h) as needed for pain. Perioperative vital signs, intraoperative time intervals, duration in the recovery and discharge times were recorded. Postoperatively, we also evaluated pain and sedation scores, total analgesic requirements, patient satisfaction, and incidence of complications (nausea, vomiting, pruritus, positional headache, backache, difficulty voiding, and dizziness). We found that in the recovery, no patients in either group asked for analgesia. VAS pain scores were very low in both groups (2.6 ± 0.8 in GA group versus 2.4 ± 1.0 in SA group after 60 min postoperatively). There were no significant differences between both groups as regards total analgesic consumption during 72h postoperatively, and sedation scores. Patients in SA group had longer time of operating room (OR) entry until starting skin preparation (13 ± 4.5 versus 5.2 ± 3.1 min) and also, from OR entry until skin incision (16 ± 5.2 versus 10.9 ± 4.7 min) in comparison with patients in GA group. However, the total duration inside OR was not different between both groups. Patients received SA had met the criteria for home readiness earlier than those received GA (68.3 ± 44 versus 95.2 ± 33 min respectively). Patients in GA group suffered more nausea than in SA (24% vs 8% respectively). The incidences of other side effects were comparable in both groups and there were no differences in patients' satisfaction scores between groups. We concluded that the two techniques with the multimodal analgesia given had provided comparable patient satisfaction and efficiencies both intraoperatively and postoperatively with low incidence of complications.