Abstract Background It is considered best practice to provide clear theoretical descriptions of how behaviour change interventions should produce changes in behaviour. Commissioners of the National Health Service Diabetes Prevention Programme (NHS-DPP) specified that the four independent provider organisations must explicitly describe the behaviour change theory underpinning their interventions. The nationally implemented programme, launched in 2016, aims to prevent progression to Type 2 diabetes in high-risk adults through changing diet and physical activity behaviours. This study aimed to: (a) develop a logic model describing how the NHS-DPP is expected to work, and (b) document the behaviour change theories underpinning providers’ NHS-DPP interventions. Methods A logic model detailing how the programme should work in changing diet and activity behaviours was extracted from information in three specification documents underpinning the NHS-DPP. To establish how each of the four providers expected their interventions to produce behavioural changes, information was extracted from their programme plans, staff training materials, and audio-recorded observations of mandatory staff training courses attended in 2018. All materials were coded using Michie and Prestwich’s Theory Coding Scheme. Results The NHS-DPP logic model included information provision to lead to behaviour change intentions, followed by a self-regulatory cycle including action planning and monitoring behaviour. None of the providers described an explicit logic model of how their programme will produce behavioural changes. Two providers stated their programmes were informed by the COM-B (Capability Opportunity Motivation – Behaviour) framework, the other two described targeting factors from multiple theories such as Self-Regulation Theory and Self-Determination Theory. All providers cited examples of proposed links between some theoretical constructs and behaviour change techniques (BCTs), but none linked all BCTs to specified constructs. Some discrepancies were noted between the theory described in providers’ programme plans and theory described in staff training. Conclusions A variety of behaviour change theories were used by each provider. This may explain the variation between providers in BCTs selected in intervention design, and the mismatch between theory described in providers’ programme plans and staff training. Without a logic model describing how they expect their interventions to work, justification for intervention contents in providers’ programmes is not clear.