Transit-Oriented Development and promoting station-area built-environment are often proposed solutions for increasing metro ridership. However, existing research has failed to consider the heterogeneity of stations and has resulted in general and ambiguous policy implications. Additionally, the high-density urban context in China raises questions about the applicability of conclusions drawn from studies conducted in Western countries. To address these issues, this study focuses on identifying station heterogeneity based on daily passenger flow distributions in Shenzhen, where 163 stations are categorized as inbound-dominated, outbound-dominated, or in-out balanced. Results from OLS regression analysis reveal that certain built-environment variables have significant effects in one category, but not in others, and even the same variable can have different coefficients across categories. These findings underscore the importance of considering station-specific factors in the planning and development of different stations, rather than relying solely on a single planning tool, such as increasing density and encouraging mixed-use development.