Although several studies have investigated the optimal treatment of knee dislocations, all have been composed of a small number of patients and therefore have made it difficult to draw definitive conclusions. The literature on knee dislocation was reviewed to allow a meta-analysis and determine whether operative or nonoperative treatment had better outcomes. Range of motion, flexion contracture, Lysholm score, instability, ability to return to preinjury employment, and ability to return to preinjury athletic activities were compared using statistical methods. A total of 132 knee dislocations treated surgically and 74 treated nonoperatively were included. Average range of motion was 123 degrees in the surgical group and 108 degrees in the nonoperative group (P<.001). Flexion contracture averaged 0.5 degrees for the surgical group and 3.5 degrees for the nonoperative group (P<.05). A significant difference (P<.001) also was found in the Lysholm scores, with a surgical group mean of 85.2 and a nonoperative group mean of 66.5. There was no significant difference in the ability to return to preinjury employment or athletic activity or in the amount of instability between the two groups. Surgical treatment appears to be associated with improved outcomes, although significant disability is still possible after successful surgical treatment.