To evaluate the cost and time effectiveness for different anesthesia methods when performing knee arthroscopy, this study compared three different anesthesia methods. Four hundred healthy patients scheduled for knee arthroscopy were randomized to either local anesthesia (LA) (n=200), spinal anesthesia (SA) (n=100) or general anesthesia (GA) (n=100). The LA arthroscopies were performed in a facility set up in our outpatient department without anesthesia service. The SA and GA arthroscopies were performed in our central operating department with full anesthesia service. Comparisons were made between total hospital stay, anesthesia time, surgery time, recovery time, pharmaceutical and total cost. The total hospital time was 130.4 min (SD 35.14, range 63-383) in the LA group compared to 280.4 min (SD 79.29, range 155-589) in the GA group (p<0.001). The total hospital time in the GA group was also shorter than that of 350.3 min (SD 65.37, range 198-502) in the SA group (p<0.001). The time from start of anesthesia to start of surgery was significantly longer in the LA group, 39.2 min (SD 13.13, range 17-87), compared to 20.1 min (SD 4.93, range 11-35) in the SA group and to 17.6 min (SD 4.64, range 9-44) in the GA group. There were no differences in the surgery time for the three groups. The use of LA was shown to save SEK 1011 (Swedish Crowns) per patient compared to SA and GA.