Results: Seven interventions were included. Three randomised controlled trials targeted tobacco using psychosocial interventions in primary health centres using the program theory: "Local Indigenous health workers extend and sustain the effects of conventional clinical brief intervention by engendering social and cultural resources". Four pragmatic trials of multiplecomponent, community-based interventions using controlled, semi-controlled or before-andafter designs used the program theory: "Discrete intervention components targeting locally defined substance misuse issues will activate latent capacities to create an environment that favours cessation. " Publications did not report clear effect, implementation fidelity or explicit mechanisms affecting participant thinking.
Conclusions:Rigorous intervention designs built on 'Roadmap' principles neither reduced substance use in the populations studied nor identified transferable mechanisms for behaviour change.Implications for public health: Substance misuse impacts among Indigenous Australians remain severe. Theoretical mechanisms of behaviour change may improve intervention design.