In this study, the correct classification level of whether forgiving oneself, others and the situation is abusing their partners was determined by logistic regression analysis. There are 221 young adults ranging from 19-30 in this study, which was designed in the scanning model. Heartland Forgiveness Scale and Information Form were used in the study. In the initial model of the analysis, all participants were classified in the group that exploited their partner, with a classification percentage of 62.9%. The biggest contribution to the initial model comes from the variable of forgiving others, respectively, the variables of self-forgiveness and forgiveness. Cox & Snell RSquare value for the final model was calculated as .10. This finding shows that when the predictor variables are included in the model, 10% of the predicted variable is explained. Accordingly, the model has a good fit. Of the 82 individuals who did not abuse the result model, 34 were classified correctly and 48 were incorrect, with a percentage of correct classification of 41.5%. Of 139 individuals abused, 120 were correct and 19 were incorrectly classified, with an accurate classification percentage of 86.3%. One-unit increase in the self-forgiveness variable is 7.90% in the odds of abuse [(1-.921).100]; One-unit increase in forgiveness to others causes an 8.10% [(1-.919).100] increase in exploit odds. Findings show that the variables of self-forgiveness and forgiveness for others make significant contributions to classifying individuals who abuse and do not. The variables of forgiving yourself and others increase the predictive power of the model created.