De-escalation is an important tool for preventing aggression in inpatient settings but definitions vary and there is no clear practice guideline. We aimed to identify how clinical staff define and conceptualize de-escalation, which de-escalation interventions they would use in aggressive scenarios, and their beliefs about the efficacy of de-escalation interventions. A questionnaire survey (n=72) was conducted using open and closed questions; additionally, clinical vignettes describing conflict events were presented for participants to describe their likely clinical response. Qualitative data were subject to thematic analysis. The major themes that de-escalation encompassed were communication, tactics, de-escalator qualities, assessment and risk, getting help, and containment measures. Different types of aggression were met with different interventions. Half of participants erroneously identified p.r.n. medication as a de-escalation intervention, and 15% wrongly stated that seclusion, restraint, and emergency i.m. medication could be de-escalation interventions. Those interventions seen as most effective were the most commonly used. Clinical staff's views about de-escalation, and their de-escalation practice, may differ from optimal practice. Use of containment measures and p.r.n. medication where de-escalation is more appropriate could have a negative impact; work is needed to promote understanding and use of appropriate de-escalation interventions based on a clear guideline.