BACKGROUND: Antimicrobial stewardship (AMS) programmes include actions to improve antibiotic use. OBJECTIVE: To identify factors of AMS interventions associated with behaviour change towards antibiotic use in hospitals, applying behavioural sciences. DATA SOURCES: The following online databases were searched: PubMed and Scopus. STUDY ELIGIBILITY CRITERIA: Studies published between January 2015 and December 2020. Outcomes required were: effect of the intervention reported in terms of antibiotic consumption, antibiotic costs, appropriateness of prescription, duration of therapy, proportion of patients treated with antibiotics, or time to appropriate antibiotic therapy. PARTICIPANTS: Healthcare professionals involved in antibiotic prescription and use in hospitals, and patients receiving or susceptible to receive antibiotics. INTERVENTIONS: Studies investigating AMS interventions in hospitals. ASSESSMENT OF RISK OF BIAS: Risk of bias was determined using the "Integrated quality criteria for review of multiple study designs" (ICROMS) tool. METHODS OF DATA SYNTHESIS: Systematic review of AMS interventions, using "Behaviour Change Wheel" to identify behaviour changes functions of interventions, and "Action, actor, context, target, time (AACTT)" framework to describe how they are implemented. Relationships between intervention functions and AACTT domains were explored to deduce factors for optimal implementation. RESULTS: Among the 124 studies reporting 123 interventions, 64% were effective in reducing antibiotic use or improving quality of antibiotic prescription. 91% of the studies had high risk of bias. Main functions retrieved in effective interventions were enablement, environmental restructuring, and education. The most common subcategories were audit and feedback and real-time recommendation for enablement function; and material resources, human resources, and new task for environmental restructuring function. Most AMS interventions focused on prescription, targeted prescribers, and were implemented by pharmacists, infectious diseases specialists, and microbiologists. Interventions focusing on specific clinical situation were effective in 70% of cases. CONCLUSION: Knowledge of factors associated with behaviour changes will help addressing local barriers and enablers before implementing interventions. The review protocol was registered in the International prospective register of systematic reviews (PROSPERO CRD42021243939).