Background. Improvement of angina pectoris symptoms after cholesterol lowering has raised questions as to the underlying mechanisms. Methods. Rabbit experiment: We compared arterial blood samples from New Zealand White cholesterol-supplemented rabbits (n = 6) with nonsupplemented rabbit samples (n = 4) in a closed-loop circulation diffusion system. The pH and partial pressures of oxygen (pO(2)) and carbon dioxide (pCO(2)) were measured continuously. The samples were first oxygen (O-2) saturated (PO2, 160 mm Hg; pCO(2), 4 mm Hg) and then desaturated in 100% nitrogen. Cholesterol levels were determined in whole blood, plasma (P Chol), red blood cells (RBCs), and RBC membranes. Human experiment: We exposed quadruple desaturated venous blood samples (n = 4) with P Chol levels of 87 to 400 mg/dL in a gas exchanger to capillary gas conditions (pO(2), 23 mm Hg; pCO(2), 46 mm Hg). After 15 minutes we performed blood gas analyses and compared our results to baseline values. Results. In the rabbit experiment the cholesterol-supplemented group as compared to the control group showed higher plasma pO(2) levels during the saturation phase and lower plasma pO(2) levels during the desaturation phase. It also had a markedly increased RBC membrane cholesterol content: 121 +/- 3 (standard error of the mean [SEM]) mg/dL versus 22 +/- 1.7 mg/dL in the control group (P < .05). This barrier to RBC membrane O-2 diffusion caused delayed O-2 entry into the RBCs during saturation with a higher plasma pO(2), and delayed O-2 release from the RBCs during desaturation, with a lower plasma pO(2). In the human experiment the P Chol level was inversely correlated with the percentage change of O-2 content in milliliters of O-2 per deciliter of blood (P < .05). Conclusions. Increased RSC membrane cholesterol in hypercholesterolemia appears to decrease the transmembrane O-2 diffusion rate.