Of the 136 wards participating in the current study, about one-third (30%) operated with their ward exit door permanently locked, whereas another third (34%) never locked the ward door. Locking the ward door only seems to have modest effects on preventing patients from leaving the ward without permission. For this reason, alternative measures for door locking, such as facilitating social contact for those at high risk of absconding, via phone contact, encouraging visits or using all available resources to enable supervised temporary leave, are recommended to prevent patients from absconding. Locking the exit doors of psychiatric wards is believed to reduce the risk of patients absconding. The aims of the study were to investigate both the prevalence of door locking and other exit security measures on UK admission wards, as well as whether door locking appears to be effective in keeping inpatients in. A cross-sectional survey on 136 acute psychiatric wards in the UK was conducted, in which a range of data on patients, staff, and conflict and containment events, including door locking and absconding, were collected from shift to shift during a period of 6 months. About one-third of the participating wards (30%) operated with their ward exit door permanently locked, whereas another third (34%) never locked the ward door. Univariate analyses suggested little association between exit security measures and absconding. A more robust multilevel statistical analysis, however, did indicate a reduction of about 30% of absconding rates when the ward door was locked the entire shift. Although locking the ward door does seem to reduce absconding to a certain extent, it far from completely prevents it. As it may be unrealistic to strive for a 100% absconding-proof ward, alternative measures for door locking to prevent absconding are discussed.