We prospectively studied 69 consecutive patients hospitalized with a primary diagnosis of acute left ventricular failure so as to assess the impact of vasodilators on incidence and morbidity of acute symptomatic left ventricular failure. The determinants of duration of hospitalization, in-hospital mortality and symptomatic status 2 months after discharge were examined. There were 9 in-hospital deaths (13%), and survival at 60 days was 77%. Median duration of hospitalization was 9 days, and 33% of the surviving patients remained in New York Heart Association functional class III-IV 60 days subsequent to discharge. Of the patients, 49 (76%) had previously received treatment for left ventricular failure: 30 (61%) of these had received vasodilators, most commonly angiotensin converting enzyme inhibitors and nitrates. Ischaemic chest pain was present in 34 (49%) of the patients. Acute utilization of vasodilators (45% of patients) was largely limited to nitrate therapy associated with ischaemic chest pain (P less than 0.01). Multiple logistic regression revealed previous left ventricular failure, advanced age and hypokalaemia as significant correlates of prolonged hospitalization (greater than 9 days). Previous left ventricular failure was also predictive of persistent severe disability two months subsequent to discharge. No factor was a significant predictor of in-hospital death. Although preceding treatment with digoxin and incremental angiotensin converting enzyme inhibitor therapy tended to predict brief hospitalization, the parameter of acute ischaemia, other biochemical anomalies and modes of acute or chronic therapy were not significant correlates of any end point. We conclude that preceding disability, rather than mode of treatment, predicts an adverse outcome in acute left ventricular failure.