Pollution exhibits significant variations horizontally and vertically within cities; therefore, the size and three-dimension spatial distribution of population is a significant determinant of urban health. This paper presents a novel methodology, three-Dimensional dIgital Geography (3DIG) methodology, for investigating both horizontal and vertical distributions of population under traffic impacts. 3DIG integrates application of Geographic Information System (GIS) and fine-resolution Digital Terrain Models (DTM). 5 meter resolution DTMs were employed to obtain the building stories in residential areas of Taipei metropolis; the vertical distribution of population at different stories were obtained accordingly. In addition, the potentially affected populations at different floors within five specified buffer zones under traffic impacts were estimated with GIS. Field survey validation indicated model results were reliable and accurate. The results showed that 97.9% (6.4 million) of the Taipei population lives within 50 m and 0.8 million (12.3%) residents live on the first or second floor within 5 m from municipal roads. This study demonstrates that 3DIG is a versatile methodology capable of identifying potentially high-exposure population and useful for various researches and policy planning in which the three-dimensional spatial population distribution is the central focus.