Seclusion and mechanical restraint are restrictive interventions that should be used only as a last resort and for the shortest possible time, yet little is known about duration of use in the broader context. Adult area mental health services throughout Victoria, Australia, were asked to complete a report form for prolonged episodes of seclusion (>8 hours) and mechanical restraint (>1 hour). The present, retrospective cohort study aimed to understand the individual (age, sex, type of service, duration of intervention) and contextual factors associated with prolonged use of restrictive interventions. Contextual factors describing the reasons for prolonged use of the restrictive interventions were captured qualitatively, and then coded using content analysis. Median duration was compared across individual factors using Mann-Whitney U-tests. During 2014, 690 episodes of prolonged restrictive intervention involving 311 consumers were reported. Close to half (n = 320, 46%) involved mechanical restraint. Seclusion episodes (n = 370) were longer in forensic mental health services compared to adult area mental health services (median: 24 hours and 18 min vs 16 hours and 42 min, P < 0.001). Mechanical restraint episodes (n = 320) were shorter in forensic mental health services compared to adult area mental health services (median: 3 hours and 25 min vs 4 hours and 15 min, P = 0.008). Some consumers were subject to multiple episodes of prolonged seclusion (55/206, 27%) and/or prolonged mechanical restraint (31/131, 24%). The most commonly occurring contextual factor for prolonged restrictive interventions was 'risk of harm to others'. Means for reducing the use of prolonged restrictive interventions are discussed in light of the findings.