By means of on-site and network investigation, we collected data relevant to residents of communities, point of interest (POI) data, and land-use data of Fuzhou. We set traffic walking time and leisure walking time as an independent variable, built environment as dependent variable, and gender, age, education level and income level as control variables. Six linear regression models were established using Statistical Product and Service Solutions (SPSS). The results showed that in the 5D (i.e., Density, Diversity, Design, Destination and Distance) elements of the built environment, the density was negatively correlated with the traffic walking time, whereas other elements were positively correlated with the walking time, but the degree of influence was different.