Abstract
Urban colour is increasingly recognised as a performative design variable influencing safety, thermal comfort, wayfinding, and energy demand, yet empirical evidence remains fragmented. This review synthesizes peer-reviewed studies using structural equation modelling (SEM) to relate objectively measured chromatic metrics to psychosocial and environmental constructs in urban contexts across China and South Korea. Focusing on exterior colour, SEM application, and model fit, 73 studies were systematically coded for metrics, constructs, quality, and contextual moderators. Three consistent causal pathways emerge: (1) mid-range colour complexity enhances perceived safety and comfort (median β ≈ 0.43); (2) dominant hue and chroma improve wayfinding and place identity (β ≈ 0.30); and (3) cool or reflective palettes reduce heat stress and annual building energy use by 2-4%. Integration of AI-driven image segmentation, extended reality (XR) user testing, and multi-level SEM frameworks has enabled more robust causal inference. Recent findings also emphasize the moderating effects of policy, density, and demographics. This review highlights how advanced spatial information technologies, digital twins, geo-computation, and real-time SEM dashboards can inform evidence-based strategies for enhancing the quality, safety, and comfort of urban living spaces. Collectively, these innovations offer transformative potential to optimize public design and foster sustainable, data-driven urban well-being.