Continuous positive airway pressure (CPAP) has been used to increase cardiac index (CI) in patients with congestive cardiomyopathy in the presence of elevated pulmonary wedge pressure. We hypothesized that with normovolemia, CPAP would decrease CI because of decreased left ventricular (LV) preload, whereas in hypervolemia CPAP would increase CI because of a decrease in afterload. We tested this hypothesis on nine sedated, unanesthetized pigs instrumented 5 to 10 d before study. We measured CI, heart rate, stroke volume, LV end-diastolic and end-systolic pressures, and LV dimensions at CPAP levels 0, 5, 10, 15, and 20 cm H2O before and after volume expansion with hetastarch (35 ml/kg). From LV dimensions, LV end-diastolic (LVEDV) and LV end-systolic volumes (LVESV) and LV ejection fraction (LVEF) were calculated. With normovolemia, CI and LVEDV decreased with increased CPAP. Volume infusion produced mild cardiac dysfunction as evidenced by increased LV volumes, decreased LVEF, and decreased contractility. With hypervolemia, CPAP produced an increase in CI, decrease in LVEDV and LVESV, and an increase in LVEF. At higher values of CPAP, we observed decreased CI and LVEDV as with normovolemia. We conclude that with normovolemia, CPAP's effects are mainly related to changes in preload. Hypervolemia produced mild cardiac dysfunction. The improvement in CI with CPAP appears predominantly to be secondary to decrease in LV afterload, but a mild preload effect, which parallels the effect seen with normovolemia, was superimposed on afterload changes at higher CPAP values.