This study examined patients' preferences for coercive methods and the extent to which patients' choices were determined by previous experience, demographic, clinical and intervention-setting variables. Before discharge from closed psychiatric units, 161 adult patients completed a questionnaire. The association between patients' preferences and the underlying variables was analyzed using logistic regression. We found that patients' preferences were mainly defined by earlier experiences: patients without coercive experiences or who had had experienced seclusion and forced medication, favoured forced medication. Those who had been secluded preferred seclusion in future emergencies, but only if they approved its duration. This suggests that seclusion, if it does not last too long, does not have to be abandoned from psychiatric practices. In an emergency, however, most patients prefer to be medicated. Our findings show that patients' preferences cannot guide the establishment of international uniform methods for managing violent behaviour. Therefore patients' individual choices should be considered.