We studied 23 consecutive patients with acute hyperventilation presenting to an inner-city emergency department, diagnosed on clinical grounds by the attending physician and confirmed by arterial blood gas values in 5 patients. An organic basis for the presenting complaints was excluded and chest radiograph, serum biochemistry, blood cell count, and thyroid function test results were normal. The male to female ratio was 12:11. Presenting complaints were dyspnea (61%), paresthesia (35%), chest pain or tightness (43%), muscle spasm (9%), dizziness (13%), palpitations (13%), and panic (30%). Similar previous episodes were reported in 74%. Misattribution of the presenting complaints to a cardiac or other life-threatening disorder was reported in 20 patients (87%) and was the main reason for their presentation to the hospital. Although no patients presented with clinical features of asthma, 7 (30%) were known asthmatics receiving treatment and another 10 (44%) had a history and investigation results suggestive of asthma. Only 2 had a history of anxiety or depression, but 17 (78%) patients exceeded the threshold for anxiety or panic on Clinical Interview Schedule (CIS-R) interview (score > or = 12). Marihuana or alcohol abuse were involved in 17% with a history of past abuse in 26%. When assessed 2 months after the attack, 13 (57%) had resting or stressor-induced hyperventilation with a significant (p < 0.05) association with asthma but not with a positive CIS-R score. These results illustrate the multifactorial basis of acute hyperventilation, the importance of misattribution, and the danger of using the term "hyperventilation syndrome" in the emergency department.