To test our hypothesis that abnormal levels of endogenous carbon monoxide (CO) produced naturally by heme oxygenase-1 (HO-1) in response to infections of all kinds may be contributing to the morbidity and mortality associated with COVID-19, we searched PubMed for peer-reviewed literature on carbon monoxide and each of eleven abnormal blood tests, fourteen signs and symptoms, and five fatal complications of COVID-19 infection reported in a case series from a hospital in Wuhan, China: acute respiratory distress syndrome (ARDS), acute kidney injury, acute cardiac injury, arrhythmia, and shock. We found reports of acute exogenous CO poisoning causing all the same signs, symptoms and complications, and all the abnormal blood tests except D-dimer and procalcitonin. Our search also found endogenous HO-1 and CO levels correlated with these complications, independent of any inhaled CO exposure. In sharp contrast to the CO poisoning literature, most studies of endogenous CO interpret its close positive correlation with these acute conditions as protective, with some going so far as to recommend treating ARDS with inhaled CO. We conclude with new recommendations for testing endogenous CO poisoning in COVID-19 cases using devices approved by the US Food and Drug Administration that can distinguish CO coming from the lungs, arteries, veins, and average of all tissues, unlike current protocols for CO poisoning that only measure CO in arteries or veins but not both. Based on these findings, we appeal to clinicians to start testing CO levels in COVID-19 patients and to stop monitoring oxygen saturation with conventional pulse oximeters that overestimate oxygen saturation by the sum of carboxyhemoglobin and methemoglobin. We conclude by reviewing FDA-approved treatments that may help COVID-19 patients with endogenous CO poisoning. These include zinc-based drugs that lower the rate of endogenous CO production by inhibiting HO-1, and drug-free devices and methods that reduce the total body burden of CO after exogenous CO poisoning.