Background Delayed discharge from hospital creates additional pressure on staff and finances. While legal sanctions are in place for acute trusts, the concept of a delayed transfer is generally poorly defined in mental health. Aims To evaluate whether levels of, and reasons for, delayed discharge have changed over a ten-year period. Method Using alternative clinical and social criteria for delayed discharge, a census survey of general adult psychiatry inpatients was taken in June 2011 and compared with a similar survey carried out in October 2001. Results The 2011 sample showed 14 per cent of clients experienced delayed discharge, reduced from 25 per cent ten years earlier. In 2011 there were fewer people awaiting rehabilitation, forensic or respite care, but more waiting to be rehoused. The proportion of new long-stay inpatients was higher in 2011, although several were in hospital for clinical reasons and were not delayed discharges. Conclusion Hospital admission appears to be used for focused treatment purposes. The inconsistent definition and application of 'delayed discharge' can lead to financial consequences for mental health services.