Heavy dependency on car use leads to traffic congestion, pollution, and physical inactivity, which impose high direct and indirect costs on society. Promoting walking and cycling has been recognised as one of the means of mitigating such negative effects. Various approaches have been taken to enhance walking and cycling levels and to reduce the use of automobiles. This paper examines the effectiveness of infrastructure interventions in promoting walking and cycling for transport. Two related sets of panel data, covering elapsed time periods of one and two years, were analysed to track changes in travel behaviour following provision of new walking and cycling infrastructure so that modal shift from private car use to walking and cycling can be investigated. Two types of exposure measures were tested: distance from the infrastructure (a measure of potential usage), and actual usage of the infrastructure. Only the latter measure was statistically significantly associated with modal shift. This in turn suggested that infrastructure provision was not a sufficient condition for modal shift, but may have been a necessary condition. Along with the use of new infrastructure, the loss of employment, higher education, being male and being part of the ethnic majority were consistently found to be significantly and positively associated with modal shift towards walking and cycling. The findings of this study support the construction of walking and cycling routes, but also suggest that such infrastructure alone may not be enough to promote active travel.