Objectives: The main objective of this research was to assess the roles of demographic and mental health variables as predictive variables of abuse among in-dividuals referred to police stations and family courts. Method:  450 individuals were selected through stratified convenience sampling from 49 police stations and the family court in the city of Tehran.  The subjects were administered a demographics questionnaire, Spouse abuse questionnaire, and the Symptom Checklist -90- Revised (SCL-90-R). The data were analyzed through multivariate regression analysis and regression variance analysis. Results: The results revealed that the variables like pregnancy, low age and education, high number of children, addiction, spouse unemployment, and psychiatric problems could be considered as predictive variables for experience of various types of abuse (physical, sexual, and emotional), (P<0.01). In addition, pregnancy, low age and education, wife's unemployment, and high number of children, are predictive variables for severe spouse abuse.  Conclusion: Identifying effective factors leading to spouse abuse can be helpful to develop purposeful programs for reducing and preventing spouse abuse.