Controversies over operative versus closed immobilization of traumatic complex multiple ligamentous knee injury are still debated. The aim of our study is to evaluate the outcome of reconstructive vs non-reconstructive (closed immobilization) treatment outcomes. This is a retrospective review of cases seen at our institution. All cases admitted with a diagnosis of knee dislocation, defined as patients with multidirectional knee instability in the setting of trauma, were reviewed. Twenty-nine consecutive patients from January of 1996 to June of 2002 were reviewed. Twenty-six patients (89.7%) were successfully recalled and their functional outcome analyzed. Comparing the operated group (n=15, 57.7%) with closed immobilized patients (n=11, 42.3%), there was no statistical difference in the range of motion (mean difference 8.55 degrees , p=0.202). While the operated group had more flexion contracture (mean difference 3.9 degrees , p=0.002), they had better stability and better overall knee function as measured by the International Knee Documentation Committee (IKDC) score (the mean difference of IKDC score was 12.13, p=0.005). In the operated group of 15 patients, we compared partial repair (n=7) with complete repair of all torn ligaments (n=8). Superior results were noted in the group with complete repair of all structures, with comparable range of motion (mean difference 0.6 degrees , p=0.861) and flexion contractures (mean difference 1.0 degrees , p=0.795) but better stability and IKDC score (mean difference of IKDC score 13.6, p=0.003). Our conclusion is that operative treatment with complete repair of all torn structures produces the best overall knee function with better knee stability and patient satisfaction.