Circulatory, chemical, and ventilatory response factors in Cheyne-Stokes respiration (CSR) were studied by experimental procedures and by mathematical analysis. Hemodynamic, ventilatory, and blood gas patterns in 16 patients with heart disease and CSR (group A) and 16 patients with congestive heart failure (CHF) without CSR (group B) were compared. CO2 and oxygen were administered to group A. The phenomena exhibited by patients with CSR were simulated by a mathematical model.The following conclusions were drawn: (1) Reduced blood flow and respiratory alkalosis were similar in both groups; (2) circulation times, lung to artery, were 29 sec greater than normal in patients with CSR but only 8 sec greater than normal in group B patients; (3) the abnormal CO2 response in CSR, a reduction in threshold and sensitivity, was similar to that in CHF, suggesting that abnormal respiratory mechanics were responsible; (4) mathematical simulation of CSR was possible by appropriate prolongation of circulation time alone; the CO2 response (reduction in threshold and sensitivity) had no significant effect on stability; (5) damping of oscillatory ventilation by CO2 inhalation could be efficiently simulated; (6) CSR with cycle periods less than 50 sec suggests increased neural or chemoreceptor excitability, anemia, or hypoxemia as contributory factors, whereas longer periods suggest primary circulatory factors.