To determine the short-term functional and medical outcomes and predictors of outcome following discharge from an acute hospital emergency department, 100 elderly (greater than or equal to 65 yr) and 100 nonelderly (less than 65 yr) patients were studied prospectively. Patients were interviewed at three days and again at three weeks following emergency department discharge. The number of new prescriptions given to both groups in the emergency department was similar (elderly 41%; nonelderly, 31%). The elderly were as likely as the nonelderly to know the correct name (elderly, 88%; nonelderly, 87%), dosage schedule (elderly, 90%; nonelderly, 90%) and purpose (elderly, 85%; nonelderly, 94%) of their new medications. There was no difference in patients' understanding of the diagnosis (elderly, 72%; nonelderly, 72%) or in medication compliance (elderly, 81%; nonelderly, 74%). Elderly patients were more likely to keep scheduled follow-up appointments (87% vs 65%; P less than .05). Despite these similarities the elderly had worse medical outcomes at three weeks; 67% of the elderly were better and 20% were worse, including seven patients who required interim hospitalization, four of whom died. In contrast, 82% of the nonelderly were better and only 4% were worse (P less than .01). None of the nonelderly required hospitalization or had died. Functional impairments were more common in the elderly both at baseline (elderly, 26%; nonelderly, 6%; P less than .01) and at three weeks (elderly, 27%; nonelderly, 5%; P less than .001). Independent predictors of poor medical outcome included age greater than or equal to 65 (P less than .009) and functional impairment at baseline (P less than .022).(ABSTRACT TRUNCATED AT 250 WORDS)