Hypoxemia and decreased intrathoracic pressure have been postulated as contributing causes of cardiovascular morbidity in obstructive sleep apnea syndrome (OSAS). Because of the difficulty of manipulating experimental conditions in humans, we developed an anesthetized closed-chest dog model, simulating the periodic airway occlusions of OSAS by periodic occlusions of the endotracheal tube (PUO). Using a periodicity of 60 s occluded, followed by 60 s ventilation for five to seven cycles, we measured heart rate (HR), cardiac output (CO), arterial pressure (Pa); left ventricular (LV) end-diastolic and end-systolic transmural pressure; dp/dt of LV pressure; left anterior descending (LAD) coronary blood flow (CBF), and regional myocardial contractility and intramyocardial pH. Four experimental conditions were studied: room air (RA) breathing (PO2 = 40); 100% O2 breathing (O2), and RA and O2 breathing with critical LAD stenosis (CS). Under all conditions PUO produced decreases in CO (10 to 30%) and proportional decreases in Pa. HR decreased, and in all but RA conditions stroke volume was unchanged. During the obstructed phase, indices of LV preload decreased. Indices of LV afterload also decreased except for LAD-perfused myocardium under RACS conditions. This latter was shown to be associated with regional ischemia (decreased regional pH and shortening). Regional ischemia was also demonstrated in two of nine dogs even under O2CS conditions. Among our major conclusions: (1) decreased Pa during PUO is due to decreased CO; (2) LV afterload does not increase during PUO; (3) with limited coronary flow reserve (CS), PUO can lead to myocardial ischemia. This is mostly but not solely due to hypoxia.