Much interest has been generated in the use of remotely sensed variables to predict the surface energy balance and the partition of net radiation into sensible, latent and conductive heat components. In urban areas such efforts are complicated by the complex three-dimensional structure of the urban surface. In this study, thermal infrared data from different observation platforms is used to model sensible heat flux for light-industrial and residential study areas within the city of Vancouver, B.C. The light industrial site is characterized by one- and two-storey warehouses with flat roofs and workshops. The residential site is composed of one- and two-storey single family dwellings. Directly measured surface energy balance fluxes are used to evaluate the estimates. Attention is directed to (1) the importance of considering the complete surface temperatures that take into account both the horizontal and vertical surfaces in urban areas; (2) to appropriately matching these with the source areas of the ground-based surface energy balance measurements; (3) to the significance of using surface rather than aerodynamic temperatures and the role of surface emissivity on this issue; and (4) the sensitivity to other data inputs and model assumptions.