The decision to use coercive measures (restraint, seclusion and forced medication) in psychiatric practice is controversial in mental health care. The EUNOMIA study was funded by the European Commission and carried out in 11 countries in order to develop European recommendations for good clinical practice on the use of coercive measures. The aim of the study is to identify sociodemographic and clinical predictors of the levels of perceived coercion in a sample of Italian patients with severe mental disorders at hospital admission. A total of 294 patients were recruited in five Italian psychiatric hospitals and screened with the MacArthur Perceived Coercion Scale to explore the levels of perceived coercion. Patients were assessed three times: within the first seven days after admission as well as after 1 and 3 months. At each time point, data on changes of perceived coercion, assessed by the Cantril Ladder of Perceived Coercion Scale, information on coercive measures received during hospitalization and the levels of satisfaction with the received treatments were collected. According to the multivariable regression model, being compulsorily admitted (OR: 2.5; 95% CI: 1.3-3.3, p < .000), being male (OR: 0.7; 95% CI: 0.9-1.4; p < .01), being older (OR: 0.03; 95% CI: 0.01-0.06) and less satisfied with received treatments (OR: -0.2; 95% CI: -0.3 to -0.1; p < .05) are all associated with higher levels of perceived coercion, even after controlling for the use of any coercive measure during hospitalization. Satisfaction with received treatment predicts the levels of perceived coercion and this should represent an important challenge for mental health professionals.