We build a conceptual model, with the legal assignment of treatment and control groups and random distribution of confounders (LARD principle) to demonstrate how urban planning and design knowledge can help discover strong natural experiments to reinforce the causal inference in healthy cities research.