Background In pregnancy, eating well, keeping active and avoiding excessive weight gain are associated with better maternal and foetal health outcomes. Dietary and physical activity (PA) interventions can be effective in changing behaviours and managing weight gain. The comparatively lower cost and greater accessibility of digital interventions make them an attractive alternative to in-person interventions. Baby Buddy is a free pregnancy and parenting app from UK charity Best Beginnings. It is designed to support parents, improve health outcomes and reduce inequalities and is actively used within the UK’s National Health Service. As such, it offers an ideal platform for delivering and evaluating a new antenatal dietary and PA intervention. Objective The aim of this four-phased study was to create a theory-based intervention within Baby Buddy to empower, encourage and support expectant parents to develop healthier dietary and PA habits for pregnancy and parenthood. Methods The intervention was developed using the person-based approach. A systematic review and meta-analysis of digital interventions targeting diet and/or PA in pregnancy (n=11) informed the rudimentary intervention concept by identifying effective behaviour change techniques. Three stages of qualitative research with pregnant and recently pregnant parents guided the intervention design. Study 1 (n=30), comprising 4 online focus groups and 12 telephone interviews, gauged response to the rudimentary concept and generated ideas for its development. Results were analysed thematically. At this stage the ‘Guiding Principles’ for the intervention development were established and regular team meetings ensured that the intervention design remained aligned with Best Beginnings’ objectives, evidence-based approach and feasibility criteria. Study 2 (n=29), comprising online individual and couple interviews, explored design ideas using wireframes and scripts and generated iterative feedback on the intervention content, branding and tone. A ‘Table of Changes Analysis’ tracked design amendments. Study 3 (n=21) tested an app prototype using ‘think aloud’ interviews with current Baby Buddy users. Public Patient Involvement and Engagement (PPIE) (n= 18) and other expert contributors (n=14) provided ad hoc input into the research process and design development. Results Study 1 confirmed the appeal and relevance of the intervention concept and its novel approach of including partners. The identified themes underpinned the development of the intervention design. Iterative feedback from Study 2, in conjunction with PPIE and expert contributor input, helped refine the intervention design and ensure its relevance and appeal to a diverse target user group. Study 3 highlighted functionality, content and design issues with the app prototype and identified ways of improving the user-experience. Conclusions The development of this co-created, theory-based behaviour change intervention adhered to the person-based approach. Listening to target users at each stage of the design process yielded actionable insights to optimise the relevance and appeal of the intervention content, tone and branding to its target audience.