Ample research has been conducted investigating the built environment (BE) impacts on pedestrian movement (PM). A clear division is evident in the literature on BE-PM links: one group tends to use geographic measures (metric distance) of the environment to explain pedestrian behaviour; the other group uses syntactic/space syntax measures (visual distance). Many review articles have been published on the former justifying contextual validity. However, relatively little is known how the effect size, directions, and consistency of syntactic measures in explaining PM vary between different geographical contexts. This paper fills this gap through a meta-analysis of published studies on this topic. Four measures (integration, connectivity, choice, and control) were studied in a random effect model, also stratified based on their contextual variation. Results show that integration, choice, and connectivity measures positively influence PM with choice being the strongest predictor (ES=0.481, p=0.000) followed by integration (ES=0.206, p=0.000) and control being an insignificant predictor. The overall findings are echoed in both developed and developing country context suggesting their contextual validity and robust effects on PM. A great opportunity exists to transfer these research findings into practice in Australia where obesity level is high among the developed countries but urban development is still predominantly focused on Greenfield sites. A Greenfield development makes it easier to configure road networks according to the research findings to promote PM and thereby to reduce the obesity level.