The study aims to compare anti-obesity interventions in a single evidence synthesis framework. Electronic databases were searched for randomized controlled trials of orlistat, rimonabant or sibutramine reporting weight or body mass index (BMI) change from baseline at 3, 6 or 12 months. A mixed treatment comparison was used to combine direct and indirect trial evidence. Ninety-four studies involving 24,808 individuals were included; 83 trials included data on weight change and 41 on BMI change. All results are in comparison with placebo. The active drugs were all effective at reducing weight and BMI. At 3 months, orlistat reduced weight by −2.65 kg (95% credibility interval −4.00 kg, −1.31 kg). For sibutramine, 15 mg gave a greater reduction than 10 mg at 12 months, −6.35 kg versus −5.42 kg, respectively. Rimonabant reduced weight by −11.23 kg at 3 months and −4.55 kg at 12 months. Lifestyle advice alone also reduced weight at 6 and 12 months, but was less effective than the pharmacological interventions. In conclusion, modest weight reductions were seen for all pharmacological interventions. Those interventions which have now been withdrawn from use (sibutramine and rimonabant) seem to be the most effective, implying that there may be a place in clinical practice for similar drugs if side effects could be avoided.