The increased popularity of short-term psychiatric hospitalization has raised the question of what kinds of patients benefit most from this treatment. The authors examined the records of 612 patients admitted to an urban general hospital's psychiatric inpatient unit over two years; all subjects had been referred by the hospital's psychiatric emergency room. A total of 156 clinical variables and 18 demographic variables recorded on the emergency room's intake form were examined and correlated with treatment outcome in terms of length of inpatient stay and disposition on discharge, as recorded on discharge records. A total of 21 clinical variables were found to be positively or negatively related to patient discharge into the community within 21 days after admission. None of the demographic features studied were unequivocally related to that outcome measure. The authors suggest that less severely ill patients may respond better (and thus have better treatment outcomes) than more debilitated ones, or that patients who recognize and acknowledge their distress may be better candidates for crisis-oriented short-term treatment.