BackgroundAbsconding from psychiatric facilities, the aetiology and impact of which have major socio-economic implications, has a multifactorial aetiological basis. Absconding patients are at higher risk of self-harm, violence, non-adherence, relapses, substance use and negative media attention. Most health professionals associate absconding with the escape of potentially dangerous psychiatric patients. Absconding causes fear and uncertainty, and portrays psychiatric services negatively. Identification of potential absconders would assist with risk assessment and prevention.AimThe aim of this study was to formulate an absconding rate as well as a descriptive profile of absconders since the inception of democracy and deinstitutionalisation.SettingThe study was conducted at Sterkfontein Hospital, a specialised psychiatric hospital outside Johannesburg.MethodsA retrospective record review of absconders from Sterkfontein Hospital in Johannesburg over 1 year was conducted.ResultsThe absconding rate was 7.83%. The characteristics of the typical absconder included single, unemployed male, early 30s, known to psychiatric services, diagnosed with schizophrenia and co-morbid substance use. An absconder is more likely to be a forensic patient not returning from official leave of absence.ConclusionThe absconding rate has decreased to less than half that of a previous study, and is within international norms. While the descriptive profile is of limited value, it does appear that psychiatric patients are being treated in a less restrictive manner resulting in fewer absconders and a change in the method of absconding. The implications for clinical practice are firstly that a clearer definition of the term absconding is needed as this will impact risk assessment and management. It is recommended that future studies separate forensic and general populations. Lastly, the formulation and use of a risk assessment tool may be of value.