A pavement is simultaneously a material arrangement, a behavioural possibility and a public-health intervention, but the three descriptions require different evidence. Michel and colleagues review 146 European articles and reports concerning mobility infrastructure, active travel, health and co-benefits. Connections between pavement networks, destinations and walking are recurrent, as are associations between active transport and certain health outcomes. Yet a literature organised around transport can privilege movements that have a destination and can overlook the circumstances that make a route desirable, negotiable or refused. The review’s useful operation is cross-scalar: it carries knowledge from street configuration through reported mobility to health and collective costs without reducing each level to the preceding one. Its policy relevance also requires restraint. A favourable modelled effect is not the same as the experience of a newly built path, and associations drawn from heterogeneous interventions are not guarantees for a particular neighbourhood. Transport planning, primary care and health assessment intersect here around infrastructure as a proposal whose consequences must still be observed.