Introduction Products of unhealthy commodity industries (UCIs) are major drivers of non-communicable diseases. The role of evidence in supporting regulation of these products remains underexplored. This study examined how different types of research and knowledge translation activities influenced policy change across unhealthy commodities.

Methods A study was conducted combining an online survey and three focus groups. Participants included researchers, civil servants, and advocates working on tobacco, alcohol and unhealthy food. An inductive coding framework was developed. Themes were subsequently interpreted using the policy cycle to examine research influence across stages of policymaking.

Results 27 survey and 15 focus group participants contributed to the study. Participants described the need for multiple, complementary forms of evidence to advance policy change, conceptualised as an ‘evidence jigsaw’. Different evidence types were perceived as particularly influential at different policy stages (eg, epidemiological evidence for agenda setting, modelling and international experience for policy formulation, lived experience and public opinion for policy adoption and monitoring and evaluation post-adoption to prevent policy rollback). Evidence on UCIs was identified as distinctive in cutting across all policy stages and was perceived as more developed and influential in tobacco control than in alcohol or unhealthy food policy. Effective knowledge translation was facilitated by collaboration between researchers and advocates and timely engagement with policymakers, while misalignment between research production and policy windows posed key challenges.

Conclusions Effective regulation of unhealthy commodities relies on assembling a range of evidence types aligned with policy needs, timing and context, alongside active knowledge translation. The experience of tobacco control illustrates how a more complete ‘evidence jigsaw’, including evidence on industry conduct, can support sustained policy action. Greater cross-sectoral learning and dialogue among public health actors might help accelerate progress in other sectors.