Background: The four major risk factors prioritised in WHO non-communicable disease prevention strategies are tobacco, alcohol, physical inactivity and, crucially, poor diet. Trans-fatty acid intake in most countries still exceeds the WHO target of 2g/day, mainly reflecting consumption of industrial trans-fats in junk food. Furthermore, the most effective policies for reducing trans-fats remain unclear. We therefore systematically reviewed the evidence on trans-fat policy interventions to inform future prevention strategies. We also explicitly compared “upstream” interventions covering whole populations with “downstream” interventions targeting individuals. Methods: A pre-piloted search strategy was used to systematically search six electronic databases (Cinahl, CRD, CDSR, Medline, SCI and SCOPUS) for papers evaluating the effectiveness of trans-fat interventions, with intake (g/day) as the main outcome measure. Two researchers independently screened, extracted and graded the papers for quality. Each study was then categorised on a continuum ranging from “downstream” interventions targeting individuals (dietary counselling, media campaigns or nutrition labelling) to more “upstream” structural interventions covering the entire population - reformulation, regulation, fiscal policies and comprehensive strategies involving multiple policies. A narrative synthesis was used to summarise and compare the effectiveness of different interventions. Results: After screening 996 candidate papers, a total of 22 papers were included in this systematic review: 12 empirical studies and 10 modelling studies. Quality was variable. The largest trans-fat reductions occurred in Denmark. Multiple interventions decreased intake from 4.5g/day in 1976 to 1.5g/day in 1995 and then virtually zero after a legislative ban in 2005. The US now intends to emulate this approach. No studies quantifying tax interventions were identified. Voluntary reformulation reduced trans-fat intake by 1.5g/day followed by worksite interventions (1.2g/day) and food labelling (0.8g/day). Dietary counselling targeting individuals achieved the smallest reductions (0.8g/day). Conclusions: Multi-component interventions including a legislative ban to eliminate trans-fats from foods appear the most effective strategy to minimise intake, as achieved in Denmark, and ongoing in the US. However this approach remains underused. Reformulation and other multi-component interventions can also achieve useful reductions. By contrast, more “downstream” interventions targeting settings or individuals consistently achieve much smaller reductions. This “effectiveness hierarchy”, previously observed in tobacco and alcohol control, might perhaps be considered more often when planning future prevention strategies; for instance to control other harmful nutrients such as salt or sugar.