Regular walking benefits health. Physical characteristics of neighbourhoods (PCONs) may be associated with walking. Objectives were to synthesize research on PCON-walking relationships and formulate a logic model of potential associations. A systematic review was conducted. MEDLINE, EMBASE, PsycINFO and TRID were searched for studies investigating PCON – walking associations. Studies were screened for inclusion, data were extracted, and risk of bias was assessed. A logic model was created by grouping PCONs into five constructs (buildings, routes, recreational areas, appearance, combined), grouping walking into two constructs (transportation, recreation), then connecting constructs with dashed lines to represent 10 hypothesized relationships. Effect estimates from the 13 included studies (all cross-sectional) were categorized into these relationships. 26 effect estimates were categorized into the ‘buildings’ – ‘transportation walking’ relationship, 24 of which were statistically significant (p < 0.05). 21 of 29 estimates categorized into the ‘combined’ – ‘transportation walking’ relationship were significant. Both proportions were statistically significant majorities (p < 0.0001, p = 0.0158). The logic model's 8 remaining PCON – walking relationships did not have statistically significant majorities of statistically significant estimates. Logic model revisions identified the significant relationships with solid lines. This evidence suggests buildings and combined PCONs are associated with transportation walking. Future cohort and quasi-experimental studies should focus on these relationships.