<p>Atmospheric particulate matter (PM) in local urban environments poses a major public-health risk, yet precise monitoring is often hindered by complex sources and strong spatial variability. Because plant leaves act as natural passive samplers, they offer a practical complement for assessing urban PM pollution. In this study, we quantified foliar dust retention for 25 representative greening species across 10 urban sites in Xuzhou, China. Dust retention differed significantly among the species and was associated mainly with leaf pubescence and surface roughness. Therefore, a principal component analysis (PCA)-based trait ranking was used only as an exploratory screen, identifying <i>Celtis sinensis</i>, <i>Magnolia grandiflora</i>, <i>Platanus acerifolia</i>, <i>Eriobotrya japonica</i>, <i>Phyllostachys viridis</i>, <i>Aucuba japonica</i> var. <i>variegata</i>, <i>Koelreuteria paniculata</i>, <i>Morus alba</i>, <i>Loropetalum chinense</i> var. <i>rubrum</i>, and <i>Eucommia ulmoides</i> as the highest-retention species. Furthermore, to interpret the captured PM, Scanning Electron Microscope-Energy Dispersive Spectrometer (SEM–EDS) analysis (supported by X-ray Diffraction) was used to resolve four particle types—silicate minerals, coal-derived carbonaceous particles, traffic soot aggregates, and clay/construction dust, with the results revealed strong site sensitivity. Specifically, site 6 was dominated by traffic soot; site 9 reflected soil resuspension and construction material transport; and coal-related particles were ubiquitous across sites, indicating a persistent combustion legacy. Foliar retention by shrubs and trees (&gt; 4.5&#xa0;m) qualitatively tracked mapped PM₁₀ patterns and showed height-dependent interception, supporting the use of urban vegetation as localized PM sinks and biomonitors. Quantitative calibration will require denser sampling, mobile monitoring, and model-corrected PM metrics.</p>