Heavy metal contamination in urban street dust is often highly heterogeneous, limiting the effectiveness of conventional geostatistical mapping approaches. In Mexico City, previous studies have reported very low spatial autocorrelation for key elements, making interpolation-based methods unsuitable for representing contamination patterns. This study proposes a multiscale cartographic framework to analyze and visualize heavy metal contamination in street dust from 482 sampling sites using the contamination factor (CF) and pollution load index (PLI) at three levels of spatial analysis: (i) city-scale patterns identified through hierarchical clustering of districts based on median CF values, (ii) district-scale variability assessed through statistical comparisons of PLI distributions, and (iii) site-scale identification of contamination hotspots using observed PLI values. Results revealed five contamination clusters and significant differences in pollution load among districts (Kruskal–Wallis, p < 0.05); PLI values in Xochimilco and Tláhuac are significantly lower than in Cuauhtémoc, Gustavo A. Madero, and Magdalena Contreras. Higher contamination levels were concentrated in northern and central districts, whereas lower levels predominated in the south. Site-scale analysis identified localized hotspots associated with transportation infrastructure, industrial areas, and commercial corridors, reflecting the influence of local emission sources. The results demonstrate that contamination patterns operate simultaneously at city, district, and site scales and cannot be adequately represented through interpolation alone. The proposed framework provides a practical approach for visualizing heterogeneous contamination datasets, supporting environmental decision-making, and may apply to other metropolitan regions characterized by weak spatial autocorrelation and localized pollution processes.