Natural ventilation (NV) is an efficient way of cooling buildings, and its energy saving potentials however depend on many parameters including local hourly weather and climate conditions, types of ventilations, indoor cooling loads (or heat gains), operating schedules, window types, and opening-wall ratios etc. Determination of the NV flow rate is thus challenging, although there are many empirical equations for different NV strategies, e.g. single-sided and cross-ventilation, considering different driving forces, e.g. wind, buoyancy and a mix of both. The main objectives of this study are to select coefficients for naturally ventilated buildings with a typical shape based on the existing empirical formulas, use the selected coefficients to develop a quick and relatively accurate method of evaluating NV potential for energy-saving analysis during early building design stages. By utilizing computational fluid dynamics (CFD), a series of computational simulations were conducted to calculate ventilation flow rate with different ventilation variables such as the wind incidence angle and the height of the building. Using the method developed, GIS maps for NV potentials of North America were created in a similar way as the well-known solar potential maps. These maps provide key graphical information of energy saving potentials of NV in terms of total hours, and associated energy savings suitable for NV for over 50 cities in the US and 10 cities in Canada.