The physical form of a city is one of the factors that shape the concentrations of air pollutants measured within it, yet the quantitative relationship between the intensity of development and measured concentrations remains poorly documented for Polish cities. The aim of this study was twofold: to determine the variation in mean annual concentrations of five air pollutants – benzo(a)pyrene (BaP), benzene (C₆H₆), NO₂, PM₁₀ and PM_2.5 – in Kraków in 2018–2024 in relation to reported emissions and meteorological conditions, and to characterise the structure and intensity of the city’s development as the spatial context of those concentrations. Data comprised annual concentrations from the State Environmental Monitoring network, emissions reported to the National Centre for Emissions Management, and building data from the Database of Topographic Objects for 90,563 buildings in 18 districts. Relationships and trends were assessed by Spearman rank correlation with Benjamini–Hochberg correction; development was described by the pseudo-usable floor area, that is the building footprint area multiplied by the number of storeys. In 2024 all five pollutants were lower than in 2018, the largest decrease being for BaP (−83.8%). Significant monotonic decreases, and significant positive emission–concentration relationships, were obtained for BaP (ρ = −1.00 and 0.96) and PM₁₀ (ρ = −0.96 and 0.93); none of the 15 meteorology–concentration pairs was significant. Low-rise and medium-rise buildings account for almost identical shares of the pseudo-usable floor area (43.68% and 43.51%), although building counts alone would indicate a decisive predominance of low-rise development (88.45%). Building intensity ranges from 10,019 m²/ha in Stare Miasto to 701 m²/ha in Nowa Huta. The picture of urban structure therefore depends strongly on the aggregation method, which should be taken into account wherever building indicators are used as proxies for dispersion conditions.