Understanding the relationship between the built environment and walking among older adults could offer important insights into land use and transport policies which seek to promote active ageing. However, most previous studies have explored global relationships, i.e. the effects are averaged or assumed to be constant over the region of interest. In this study, we focus on the local spatial variations in the relationship between the built environment and the daily time spent walking by older adults. We apply a geographically weighted regression (GWR) model, using data collected from 702 older adults in Nanjing, China. Our results show that spatial heterogeneity exists for built environment effects within the entire study area. It has an impact on all the relationships, with nuances in the significance level, parameter magnitude or sign reversals, depending on the location. Therefore, policy interventions would only be effective in certain areas for certain built environment attributes. By exploring the local contexts of relationships, we further suggest that the spatial heterogeneity stems from contextual effects, i.e. the specificities of places with a discriminative composition of individual and/or environmental characteristics. Our findings can help to enrich the understanding of associations between land use and travel behaviour, as well as offer local planning guidance for creating age-friendly neighbourhoods.