Evidence suggests the prevalence and duration of mechanical restraint are particularly high among forensic psychiatric inpatients. However, only sparse knowledge exists regarding the reasons for, and characteristics of, prolonged use of mechanical restraint in forensic psychiatry. This study therefore aimed to investigate prolonged episodes of mechanical restraint on forensic psychiatric inpatients. Documentary data from medical records were thematically analyzed. Results show that the reasons for prolonged episodes of mechanical restraint on forensic psychiatric inpatients can be characterized by multiple factors: "confounding" (behaviors associated with psychiatric conditions, substance abuse, medical noncompliance, etc.), "risk" (behaviors posing a risk for violence), and "alliance parameters" (qualities of the staff-patient alliance and the patients' openness to alliance with staff), altogether woven into a mechanical restraint spiral that in itself becomes a reason for prolonged mechanical restraint. The study also shows lack of consistent clinical assessment during periods of restraint. Further investigation is indicated to develop an assessment tool with the capability to reduce time spent in mechanical restraint.