Repetition of psychiatric emergency department use by a relatively small number of patients constitutes a major problem for clinicians and service providers. This study aimed at the identification of risk factors for repetition by addressing the time interval between the first and second visits to the emergency department. The purpose was to investigate what patient characteristics and referral circumstances determine this interval. Over a two year period, data on all psychiatric emergency referrals to the emergency department of four public hospitals were collected with a standardized form. Data collected during the index referral of all patients were used for estimating the risk for repetition using survival analysis techniques. A large proportion of repeaters revisits the emergency department within a short time interval. Younger, male patients who present themselves spontaneously are more likely to repeat than others. Previous inpatient service use and the presence of a diagnosis of substance abuse disorder or psychotic disorder at the first visit further increases the risk for repetition. Previous service use and, to a lesser degree, demographic and clinical characteristics of psychiatric patients are useful in the prediction of variations in time between first and second referrals to the emergency department