We retrospectively reviewed the charts of 308 admissions to a pulmonary disease ward and 100 admissions to the general medical service over one year to find the prevalence, sequelae, and etiology of hypophosphatemia. The overall prevalence of low serum phosphate levels (less than 2.4 mg/dl) occurring at least once during hospitalization in chest patients was 17 percent, but was higher in patients with respiratory infections (28 percent). Moreover, the prevalence of hypophosphatemia on admission (before institution of intravenous fluid or drug therapy) was ten times higher in patients with respiratory infections than in patients with noninfectious respiratory illness or general medical patients (21 vs 2 percent, p less than 0.001). Serum phosphate less than 2.0 mg/dl occurred in 4 percent of patients. Twenty-seven percent of the patients (including two with ventilatory failure) with abnormally low serum phosphate levels had symptoms or signs of uncertain etiology later explicable by the presence of hypophosphatemia. The most common additional laboratory finding associated with hypophosphatemia was elevation of muscle enzymes. Although mortality was no higher in hypophosphatemic patients, hospital stay was twice as long as that of patients with normal levels of serum phosphate. No correlation was found between simultaneous arterial blood gases and serum phosphate levels. Two patients given antacids had severe hypophosphatemia and worsened ventilatory function; phosphate-binding antacids should be used judiciously in patients with severe respiratory disease, since they may lead to the development or worsening of hypophosphatemia and diminished ventilatory function.