The relationships between inequality and crime and segregation and crime are classic topics in crime geography. However, the effects of city-level inequality and segregation based on an individual's socioeconomic status on crime in a racially homogeneous country such as China remain largely unexplored. Using novel location-based service population big data, criminal judgment records, and a nationally representative survey, we measure the degree of educational inequality and educational segregation in 288 Chinese cities and then simultaneously examine their effects on theft and violent crime rates. The results suggest that inequality and segregation are two distinct constructs that reflect social polarization, as evidenced by their different spatial distribution patterns and low correlations. Both educational inequality and educational segregation have significant and positive effects on the incidence of theft and violence at the city level. These relationships remain robust to introducing instrumental variables to address the endogeneity issue. Considering these findings, it is imperative to take a multi-sectoral and coordinated approach to reducing inequality and segregation and, ultimately, deterring crime. By understanding and acting upon these factors, policymakers can work toward promoting a safer and more equitable society across China.