The albedo of urban roofs and impervious surfaces constitutes one of the key parameters in the energy balance of the city environment, and its detailed and accurate assessment represents an ambitious and challenging objective. In this work, an analysis of Landsat Thematic Mapper (TM) data and airborne infrared (IR) thermography data over the city of Florence was performed to assess the satellite sensor capability to successfully detect the surface albedo with a high level of detail over a heterogeneous urban texture. The effect of the TM reflective bands spatial resolution (30 m) on the albedo retrieval of single roofs, squares or roads was investigated, comparing these data with the aircraft surface temperature measurements at 1 m spatial resolution, which are indirectly linked to the surface reflective properties. The detailed analysis of the albedo pattern from spaceborne observations provided reliable albedo values for dark and light roofs and for low albedo surfaces; different significant situations were investigated, and interesting results came up. The Landsat TM albedo succeeds on detecting dark and light roofs, as well as the high resolution IR thermography, but if the latter is averaged in a 30 m grid (the same of the albedo), the most significant information gets lost. Besides, in the presence of low albedo materials, occasionally cold surfaces appeared in the airborne image due to the shadow of buildings, a circumstance that could bring to wrong evaluations of the surface reflective properties, trusting only on thermal images. Moreover, it is shown how the different exposures to the sun of the same surface affected the thermography results, a circumstance not found on Landsat TM albedo pattern. The paper demonstrates that satellite data show an effective potential for the assessment of the albedo of roofs and impervious surfaces, even when a high level of detail is required, providing information for improving urban planning efforts to dampen heating effects.