Abstract The advance and retreat of a lake-breeze front in the urban, lakeshore environment of Toronto was observed by two instrumented towers, one in the downtown business section and the other in the suburbs. In the city, there was a greater temperature contrast between the inland air and the incoming lake air than there was over the suburbs so that the inversion accompanying the advancing front was much more intense in the city. In the suburbs, the inversion associated with the retreating front had roughly the same intensity and duration as that during the advance. By contrast, heating in the city caused the retreating front to appear only as a poorly defined period of weak lapse and isothermal conditions at the downtown tower. A simple model was used to estimate the slopes and thicknesses of the front. Results were promising but a better knowledge of the speed of travel of the front is required. Estimates of 1:11–1:61 were obtained for the slopes; the thicknesses were estimated to vary from 140 to 4000 m. Differences in slopes and thicknesses were not relatable to city/suburban contrasts in surface heating and roughness. It was found that the concentrations of suspended particulates were not affected on an hour-to-hour basis by the lake breeze, but rather by the synoptic conditions giving rise to the lake breeze; in particular, the light winds associated with a slow moving anticyclone. Similarly, the concentrations of total oxidants in the city were uneffected by the lake breeze.