The world health organization suggested that violence against women is not a new phenomenon, norrnare its consequences to womenâ€™s physical, mental and reproductive health. What is new is the growingrnrecognition that acts of violence against women are not isolated events but rather form a pattern ofrnbehavior that violates the rights of women and girls, limits their participation in society, and damagesrntheir health and well-being. Wife beating, the most widespread form of domestic violence, is a majorrncause of disability and death in many developing countries like Ethiopia. This study aims to assess thernattitudes of men toward wife beating using data from the Ethiopia demographic and health surveyrnconducted in 2016. The binary multiple logistic regression and LASSO logistic regression models wererncompared to identify the best model fit by K-fold Cross-Validation in identifying some importantrnpredictors of attitudes of men toward wife beating. The LASSO Logistic regression model fit resultsrnare preferred over the binary multiple logistic regression because of the fact that it is parsimonious withrnthe same average predicting accuracy and findings from this model showed that the predictors: region,rnplace of residence, education levels, drinking status, and age of men are significantly associated withrnmenâ€™s attitudes toward wife beating. Offering the access to education for men and raising their level ofrneducation in the whole society may help reducing wife beating.