Walking frequency is not a transparent expression of an objectively walkable neighbourhood. Liu, Helbich and Ettema compare self-reported non-commute walking among 366 adults aged 45–65 in Rotterdam and Maastricht with perceived environmental conditions, health indicators and objective neighbourhood characteristics calculated at several home-centred radii. Zero-inflated Poisson models associate walking positively with address density and perceived safety, while street connectivity and perceived aesthetics show inverse associations in this sample. Body mass index is negatively associated with reported frequency, yet neither body mass index nor self-perceived health significantly moderates the environment–walking relationships tested. The methodological operation is to place multiple environmental descriptions alongside a behavioural outcome, then test rather than assume their relative and interactive explanatory roles. It bridges transport geography, urban analytics and public health, making possible an account in which modelled morphology and perception each contribute without having to agree. The observational design and reported walking measure do not establish causal route-level mechanisms. Precisely for that reason, its results challenge efforts to read individual urban experience directly from residential buffers.
