Aim: The specifics of the potential predictive risk factors of violence and violent offending in patients with paranoid schizophrenia in Croatia have not yet been studied or fully understood. Based on the body of literature and commonly cited predictive risk factors in other countries a cross-sectional study was designed with aim to test the potential similarities and or specificities of the risk factors in the Croatian patient group. Method: The cross-sectional study of male (in)patients with paranoid schizophrenia with (N=100) or without (N=100) history of physical violence and violent offending was conducted simultaneously in several hospitals in Croatia during a one year period. The data pertaining to their sociodemographic characteristics, the duration of the untreated illness phase (DUP), alcohol abuse, suicidal behavior, personality features and the insight into the illness belonging to the both groups of patients were carefully collected by trained specialist researchers and compared by experienced biomedical statistician. Additionally, the binary logistic regression model was then used to determine potential significant predictors of violence and violent behavior of patients. Results: In this study, the duration of the untreated illness (DUP) prior to first contact with the psychiatric services and alcohol abuse were shown to be highly predictive of the violent forensic history. Of all the positive predictive factors for violence and violent offending, scoring more than 8 on the AUDIT test for alcohol abuse was found in the regression model as the strongest positive predictive factor. Additionally, psychopathy index, emotional stability, conscientiousness were also found as significant positive predictive factors. Conversely, extroversion, pleasantness and intellect were amongst significant negative predictive factors for violence and violent offending. Conclusion: The association between alcohol abuse and unrecognized symptoms of desease and the risk for violence and violent offending in schizophrenia was strongly born out by results of our study. It is hoped that this finding will further inform and enable improved public and mental health prevention strategies in this vulnerable patient group. Equally, our study highlighted the need for further prospective studies in order to investigate some potential particularities of Croatian sample (e.g. older age of the forensic patients) which were suggested by our research.