Crime is an unplanned change due to the process of urban development. The geographical approach is able to analyze spatial interactions of criminogenic factors into crime phenomenon. This study aims to explain the spatial interaction between socio-economic conditions and crime risk in Bandung City. The crime risk is obtained through Kernel Density Estimation (KDE) method based on The Annual Crime Data from Bandung Police Department, whereas the socio-economic condition data is revealed from interview and questioner with 176 respondents who determined by multistage random sampling in 37 hamlets. Variable interactions are analysed based on Geographically Weighted Regression (GWR) with Gaussian Distance Function type of cross validation as optimum weight. The results show the socio-economic conditions has a positive and significant effect on crime risk. This model shows that the independent variable acts as a crime attractor. Partially its known that economic conditions have a stronger interaction with crime risk than the social climate. The strongest interaction between variables are concentrated in the southern of Kebon Pisang Urban Village, precisely in the area passed by arterial roads at southern of Sumur Bandung. Crime prevention efforts are conducted using environmental interventions by the formal security personnels and strengthening the social climate of the community.