Local anesthesia (LA) for outpatient knee arthroscopy is not a standard procedure at most hospitals. To evaluate the LA technique for knee arthroscopy on medically healthy patients, this study compared 3 anesthesia techniques. Four hundred patients were randomized to either local (n = 200), general (n = 100), or spinal (n = 100) anesthesia. Evaluated outcomes included the patient's subjective view of the procedure, and nausea and pain at rest and during active movement. All variables were recorded perioperatively and postoperatively. In addition, the performing surgeon's opinion of the degree of patient pain and the technical difficulty of the procedure were noted. Three hundred forty-two patients completed the study. In the group receiving local anesthesia (n = 180) the median visual analog scale pain score during surgery was 6 mm (mean, 17.5; SD, 23.2; range, 0 to 100 mm). Twenty-one LA patients would have preferred another form of anesthesia. In 29 patients, LA was not considered as the optimal anesthesia by the performing surgeon. Eight LA patients agreed with the surgeon that the anesthesia method used was not optimal, of these patients, 5 had synovitis (3%). In 5% of the LA patients there were technical problems. Thus, this study shows that elective knee arthroscopy can be performed under local anesthesia in 92% of the patients from a technical point of view. Excluding patients who do not choose local anesthesia and those who have hypertrophic synovitis preoperatively, knee arthroscopies can be performed as safely and effectively under local anesthesia as under any other form of anesthesia. For most patients, local anesthesia can be recommended as the standard procedure for outpatient knee arthroscopy.