The present study examines the correlates of successful outcome (subsequent inpatient or outpatient treatment) of 305 outpatient cases treated by crisis intervention therapy. Although the two measures of outcome were correlated with each, the respective relationships with the predictor variables were sometimes different. Demographic variables and therapist's characteristics were not correlated with either outcome measures. Past responses to stress (prior state hospitalization and suicidal behavior) were predictors of subsequent inpatient treatment, but not subsequent outpatient treatment. On the other hand, current precipitating events did not predict inpatient treatment, but did predict subsequent outpatient treatment. Several statistically nonsignificant relationships which either replicate or disagree with previous findings are also discussed.