Using data from a psychiatric health centre, this thesis investigates variables affecting the revolving door syndrome. An exploratory direct effects model incorporated social psychiatric and social background variables on recidivism. Initial data analysis consisted of frequency distributions and difference of means tests between first admission patients and recidivist patients. Compared to first admission patients, revolving door patients were more likely to have an affective psychoses disorder, secondary diagnosis of drug and/or alcohol abuse, psychiatric illness in their family, to be unemployed, and to have suffered sexual abuse. The second stage of the data analysis used a multiple regression analysis that included only patients from the recidivist sample. Significant predictors of psychiatric recidivism include a patient's psychiatric history, the number of in-patient days, their General Assessment of Functioning score on admission, recent financial loss, and age. This investigation revealed important significant direct effects on psychiatric recidivism. The findings of this study suggest future research objectives as well as policy implications for mental health professionals.