BackgroundImproving transport infrastructure to support walking and cycling on the journey to and from work – active commuting – could help to promote physical activity and improve population health.AimsTo assess whether or not investment in new high-quality transport infrastructure was associated with an increase in active commuting; wider health impacts of changes in travel behaviour; determinants of the use and uptake of active commuting; and how changes in travel behaviour were distributed in the population and related to the wider social context.DesignThe Commuting and Health in Cambridge study, comprising a quasi-experimental cohort study combined with both nested and supplementary in-depth quantitative and qualitative studies.SettingCambridgeshire, UK.ParticipantsA cohort of 1143 adults living within 30 km of Cambridge, working in the city and recruited in 2009; and a separate sample of 1710 users intercepted on the Cambridgeshire Guided Busway in 2012.InterventionThe Cambridgeshire Guided Busway, comprising a new bus network using 22 km of guideway (segregated bus track) accompanied by a traffic-free path for pedestrians and cyclists, opened in 2011.Main outcome measureChange in time spent in active commuting from 2009 to 2012, using a self-reported measure validated using georeferenced combined heart rate and movement sensor data.MethodsA delay from 2009 to 2011 in completing the intervention entailed some changes to the original design and attrition of the cohort. A period of methodological and observational research on active commuting preceded the evaluation, which was based on a quasi-experimental cohort analysis together with the intercept and qualitative data. A graded measure of each participant’s exposure to the intervention, based on the proximity of the busway to his or her home, served as the basis for controlled comparisons.ResultsCommuting practices were complex and shaped by various changeable social and environmental factors. Walking and cycling were often incorporated into longer commuting journeys made predominantly by car or public transport. In multivariable multinomial regression analyses, exposure to the intervention was associated with a greater likelihood of a large increase in the proportion of commuting trips involving any active travel [adjusted relative risk ratio (RRR) 1.80, 95% confidence interval (CI) 1.27 to 2.55], of a large decrease in the proportion of trips made entirely by car (RRR 2.09, 95% CI 1.35 to 3.21), and of an increase in weekly cycle commuting time (RRR 1.34, 95% CI 1.03 to 1.76). There was a mixed pattern of effects at the individual level, with the intervention providing a more supportive environment for active commuting for some and not for others. There was some evidence that the effect was most pronounced among those who reported no active commuting at baseline, and observational evidence suggesting a relationship between active commuting, greater overall physical activity, and improved well-being and weight status.ConclusionsThese findings provide new empirical support and direction for reconfiguring transport systems to improve population health and reduce health inequalities. They should be combined with evidence from research evaluating related environmental changes in other settings, preferably using longer periods of observation and controlled comparisons, to support more generalisable causal inference.FundingThe National Institute for Health Research Public Health Research programme.