As China enters the stage of "New Normal" development, promoting people's subject well-being has become an important target of the new-type urbanization. Meanwhile, improving the living environment is considered an inevitable requirement for improving the quality of urbanization. Although a substantial body of literature has focused on the linkage between the neighborhood environment and subjective well-being, only a few studies have attempted to unravel the intermediate mechanism of the relationship. Additionally, previous research mostly evaluated neighborhood built environment based on the administrative jurisdiction, rather than the real activity space of residents, ignoring the modifiable areal unit problem (MAUP). Using the survey data collected in 23 residential communities between June and August 2015 in Guangzhou City, China, this study explored the built environment factors that affect subjective well-being with a multilevel linear regression model, paying special attention to the impact of buffer zone redrawing on the analysis results. Mediation analysis was also adopted to reveal the mediating effect of residential satisfaction, while stratified analysis was applied to explore the impact of respondents' objective socioeconomic status heterogeneity on the relationship between built environment and subjective well-being. Several conclusions were drawn as follows: 1) Compared with the indices extracted from administrative jurisdiction and 1000 m buffer, the indices extracted from 15-minute walking distance have higher correlation with residents' subjective well-being. 2) Per capita green space area and point of interest (POI) density are positively correlated with residents' subjective well-being, while population density is negatively correlated. 3) Mediation analysis indicated that residential satisfaction fully mediates the relationship between POI density and subjective well-being. 4) The correlation between neighborhood built environment and subjective well-being is significantly different due to heterogeneity of individual socioeconomic characteristics (that is, hukou status, housing tenure, and employment status). Our findings reaffirm the significant role of the real neighborhood environment in supporting subjective well-being, provide some insights for urban planners and policymakers on how to interweave well-being promotion into urban environments. It is suggested that the needs of people with different socioeconomic characteristics should be taken into account to develop targeted 15-minute community life circle planning. Additionally, strengthening green infrastructure construction also matters.