Background. Modifying transport infrastructure to support active travel (walking and cycling) could help to increase population levels of physical activity. However, there is limited evidence for the effects of interventions in this field, and to the best of our knowledge no study has convincingly demonstrated an increase in physical activity directly attributable to this type of intervention. We have therefore taken the opportunity presented by a 'natural experiment' in Cambridgeshire, UK to establish a quasi-experimental study of the effects of a major transport infrastructural intervention on travel behaviour, physical activity and related wider health impacts. Design and methods. The Commuting and Health in Cambridge study comprises three main elements: a cohort study of adults who travel to work in Cambridge, using repeated postal questionnaires and basic objective measurement of physical activity using accelerometers; in-depth quantitative studies of physical activity energy expenditure, travel and movement patterns and estimated carbon emissions using household travel diaries, combined heart rate and movement sensors and global positioning system (GPS) receivers; and a longitudinal qualitative interview study to elucidate participants' attitudes, experiences and practices and to understand how environmental and social factors interact to influence travel behaviour, for whom and in what circumstances. The impacts of a specific intervention - the opening of the Cambridgeshire Guided Busway - and of other changes in the physical environment will be examined using a controlled quasi-experimental design within the overall cohort dataset. Discussion. Addressing the unresolved research and policy questions in this area is not straightforward. The challenges include those of effectively combining different disciplinary perspectives on the research problems, developing common methodological ground in measurement and evaluation, implementing robust quantitative measurement of travel and physical activity behaviour in an unpredictable 'natural experiment' setting, defining exposure to the intervention, defining controls, and conceptualising an appropriate longitudinal analytical strategy.