Objective: Results of testing the original four-factor structure of the URICA—DV as well as two three-factor structures in prior researches is presented to investigate alternative structures suitable for Korean batterers. Method: This study developed the URICA—DV Korean Version and sought the most appropriate structure for a Korean context. The sample is 293 batterers either referred by the courts, by prosecution, or who participated voluntarily. Results: The original four-factor model and the two three-factor models are demonstrated as possessing poor model fit index and fail to be accepted as apt models for a Korean context. The suggested three-factor model is a better fit according to confirmatory factor analysis. Conclusion: Findings and further discussion for practical implications and research are presented.