BACKGROUND: Core outcome sets are standardised sets of outcomes that should be collected and reported for all clinical trials. They have been widely developed and are increasingly influential in clinical research, but despite this, their use in public health has been limited to date. The aim of this study was to develop a core outcome set for public health trials evaluating interventions to promote physical activity in the general adult population. METHODS: The core outcome set was developed using a three-stage approach: stage one: a review of literature to identify potential domains for inclusion in the core outcome set; stage two: a Delphi survey was carried out to reach consensus about which outcome domains to include in the core outcome set; and stage three: a second Delphi survey was conducted to determine how best to measure the outcome domains included in the core outcome set. RESULTS: A classification of 13 outcome domains of physical activity was developed (stage one). Twenty people completed round one of the first Delphi survey (stage two), reaching a consensus to include two domains in the core outcome set, 'device-based level of physical activity' (80.0%, n = 16) and 'health-related quality of life' (70.0%, n = 14). No further consensus on the remaining outcome domains was reached in round two. Nineteen people completed the second Delphi survey (stage three). Participants rated the accelerometer (mean rating = 3.89, on a scale of 1 (do not recommend) to 5 (highly recommend)) as the best device to measure level of physical activity, and the EQ-5D (73.7%, n = 14) as the most appropriate measure of health-related quality of life. CONCLUSIONS: This study has made progress towards the development of a core outcome set for use in physical activity trials, however, there was limited consensus about which domains to include. The development of the core outcome set was challenged by the need for trial-specific outcomes, and the complexities of collecting, processing and reporting device-based data.