BackgroundIn the UK, responsibility for many public health functions was transferred in 2013 from the National Health Service (NHS) to local government; a very different political context and one without the NHS history of policy and practice being informed by evidence-based guidelines. A problem this move presented was whether evidence-based guidelines would be seen as relevant, useful and implementable within local government. This study investigates three aspects of implementing national evidence-based recommendations for public health within a local government context: influences on implementation, how useful guidelines are perceived to be and whether the linear evidence-guidelines-practice model is considered relevant.MethodsThirty-one councillors, public health directors and deputy directors and officers and other local government employees were interviewed about their experiences implementing evidence-based guidelines. Interviews were informed and analysed using a theoretical model of behaviour (COM-B; Capability, Opportunity, Motivation–Behaviour).ResultsContextual issues such as budget, capacity and political influence were important influences on implementation. Guidelines were perceived to be of limited use, with concerns expressed about recommendations being presented in the abstract, lacking specificity and not addressing the complexity of situations or local variations. Local evidence was seen as the best starting point, rather than evidence-based guidance produced by the traditional linear ‘evidence–guidelines–practice’ model. Local evidence was used to not only provide context for recommendations but also replace recommendations when they conflicted with local evidence.ConclusionsLocal government users do not necessarily consider national guidelines to be fit for purpose at local level, with the consequence that local evidence tends to trump evidence-based guidelines. There is thus a tension between the traditional model of guideline development and the needs of public health decision-makers and practitioners working in local government. This tension needs to be addressed to facilitate implementation. One way this might be achieved, and participants supported this approach, would be to reverse or re-engineer the traditional pipeline of guideline development by starting with local need and examples of effective local practice rather than starting with evidence of effectiveness synthesised from the international scientific literature. Alternatively, and perhaps in addition, training about the relevance of research evidence should become a routine for local government staff and councillors.