The use of seclusion to manage conflict behaviours in psychiatric inpatient settings is increasingly viewed as an intervention of last resort. Many protocols have, thus, been developed to reduce the practice. We conducted a systematic review to determine the effectiveness of protocols to reduce seclusion on process outcomes (e.g., seclusion, restraint), patient outcomes (e.g., injuries, aggressive incidents, satisfaction), and staff outcomes (e.g., injuries, satisfaction). We searched Medline, Embase, the Cochrane Register of Clinical Trials, PsycINFO, CINAHL, cairn.info, and ClinicalTrials.gov for protocols to reduce seclusion practices for adult patients on inpatient mental health units (from inception to September 6, 2022). We summarised and categorised reported elements of the protocols designed to reduce seclusion using the Behaviour Change Wheel Intervention Functions and resources needed to implement the protocol in psychiatric units. We assessed risk of bias and determined certainty of evidence using GRADE. Forty‐eight reports addressed five approaches to reduce seclusion: hospital/unit restructuring (N = 4), staff education/training (N = 3), sensory modulation rooms (N = 7), risk assessment and management protocols (N = 7), and comprehensive/mixed interventions (N = 22; N = 6 without empirical data). The relationship between the various protocols and outcomes was mixed. Psychiatric units that implement architecturally positive designs, sensory rooms, the Brøset Violence Checklist, and various multi‐component comprehensive interventions may reduce seclusion events, though our certainty in these findings is low due to studies' methodological limitations. Future research and practice may benefit from standardised reporting of process and outcome measures and analyses that account for confounders.