Despite human well-being serving as a crucial indicator for measuring social development and progress, the measurement of well-being faces great challenges. This study constructs a multi-level statistical model to investigate the geographical relationship between subjective well-being measure derived from a large-scale social survey and objective well-being measure derived by using a non-compensatory indicator aggregation technique and a variety of urban geospatial data. We find that a consistent resemblance between the subjective and objective measures of well-being at both the community (r = 0.61 with a p < 0.01) and sub-district scales (r = 0.62 with a p < 0.01) within a city. This finding is further supported by the estimation results from models that controls for a variety of covariate effects. Drawing on the close resemblance between the subjective and objective measures of human well-being, we develop a statistical model that links these two measurements to create a composite measure of well-being. Our results extend the understanding on the geographical match between subjective and objective measures of well-being at an intra-city spatial scale, and provide a new insight on how to leverage the emerging urban geospatial data and social survey when devising indicators of well-being.