Many researchers have the goal of improving buildings daylight availability in order to decrease the HVAC energy intake. Indeed by achieving/identifying sustainable urban forms and their design concepts, the benefits of the solar radiation can be of threefold: 1) placing PV collectors on the roofs and/or facades produces electricity, 2) placing water heating systems on the roofs and/or facades saves fossil fuel, and 3) having well dimensioned windows diminishes the needs of artificial lighting and reduces the heating/cooling bill. The way followed until now to improve the buildings performance was to evaluate it using software like PPF (a simulation programme for predicting urban solar potential) on various urban forms, and choose the best one. A new methodology is presented in this paper; it enhances the performance of buildings without the expense of trial and error in usual simulation software. The methodology, based on evolutionary algorithms, could provide helpful insights for planning solar cities.