The aim of this study was to determine what patient characteristics are used to decide whether a patient is or is not admitted to a psychiatric hospital, and what happens to those not admitted. A further aim was to determine if high levels of risk on admission predict seclusions, length of stay, or readmission within 28 days. Data were collected prospectively on consecutive presentations to an admission office via case notes and electronic databases. Eighty percent (100/127) of the adults presenting to the admission office over a typical month were admitted to hospital. Patients were more likely to be admitted if they were experiencing psychosis or exacerbation of schizophrenia, referred by other doctors or mental health teams, had a legal reason for referral, or if they were homeless. There was no association between risk for violence or suicide and seclusion rates, length of stay, or being readmitted within 28 days. It was reassuring to find that 85% of those not admitted were referred to other mental health providers, and none required admission over the following month. This study found high rates of seclusion and readmissions within 1 year, which requires further study to find strategies to reduce these rates.