SummaryObjectives: To assess the efficacy and safety of low molecular weight heparin (LMWH) as deep venous thrombosis (DVT) prophylaxis following total knee arthroplasty. Data sources: Medline 1986 to June 1997, Embase, and manufacturers were used to identify randomized controlled trials. Review methods: Trials included were randomized studies of LMWH with routine radiological screening for DVT. Placebo or active controls were included. Two reviewers independently screened trials for inclusion, and assessed their quality. Pooled relative risk estimates of DVT and proximal DVT rates were calculated using a DerSimonian and Laird random effects model. Sensitivity of the results to the type of control used and the quality of the trial was assessed. Results: The relative risk of DVT for a patient given LMWH is 0.73 (95% CI 0.66 to 0.80) when compared with patients treated with adjusted dose heparin or warfarin controls. The relative risk for proximal DVT is 0.58 (95% CI 0.38 to 0.90). The relative risk of pulmonary emboli in the LMWH group was 0.55 (95% C.I. 0.20 to 1.57). No excess of bleeding was recorded in the LMWH group. Conclusions: Low molecular weight heparin is more efficacious than either adjusted dose heparin or adjusted dose warfarin, when used to prevent DVT and proximal DVT following total knee arthroplasty.