As residential patterns and types are closely related to population densities, carbon emissions from residents’ transportation are logically correlated with the population density, as well as the diversity of facilities and the accessibility of transportation stations. This paper takes Shanghai Caoyang Xincun as a case, and studies the correlation between the residential patterns and carbon emissions through questionnaire data analysis on eight types of residential patterns. The results show that the relationship between the population density of Caoyang Xincun and the per capita traffic carbon emissions is a nonlinear fl uctuation, that the relationship between the diversity of facilities and carbon emissions is positively correlated, and that the relationship between the accessibility of transportation stations and carbon emissions is negatively correlated. It also finds that the per capita carbon emissions in residential areas with multi-storey enclosed buildings and small high-rise row buildings are different from other patterns. With a further study on occupational structure and family income, it concludes that the carbon emission difference is caused by consumer segregation based on social structure. Therefore, when judging whether a residential pattern is low carbon or not, the spatial and social implications of density should be clarified first. Then the national and regional development conditions need to be considered to guide the future development of residential patterns. As residential patterns and types are closely related to population densities, carbon emissions from residents’ transportation are logically correlated with the population density, as well as the diversity of facilities and the accessibility of transportation stations. This paper takes Shanghai Caoyang Xincun as a case, Audience studies the correlation between the residential patterns and carbon emissions through questionnaire data analysis on eight types of residential patterns. The results show that the relationship between the population density of Caoyang Xincun and the per capita traffic carbon emissions is a nonlinear fl uctuation, that the The relationship between the diversity of facilities and carbon emissions is positively correlated, and that the relationship between the accessibility of transportation stations and carbon emissions is negatively correlated. It also finds that the per capita carbon emissions in residential areas with multi-storey enclosed buildings and small High-rise Row buildings are different from other patterns. With a further study on occupational structure and family income, it concludes that the carbon emission difference is caused by consumer segregation based on social structure. The spatial and social implications of density should be clarified first. Then the national and regional development conditions need to be considered to guide the future development of residential patterns.