What level of carbon monoxide is considered dangerous?

I downvoted Jeff's answer [to his own question] because the table he posted without comment that specifies CO symptoms supposedly associated with various CO concentrations--although correctly referenced to its immediate source [a company called Nyad that makes process gas analyzers]--is unreferenced by Nyad, apocryphal, and unsupported by any actual medical studies. As discussed in my edit of the question, it is meaningless to speculate about how dangerous the inhalation of any given concentration of CO may be without also specifying the duration of exposure and the minute ventilation rate during exposure.  Tables that do not specify these variables are thus quite misleading and should not be cited as authoritative..  There are hundreds of peer-reviewed medical studies on CO poisoning that have noted how prolonged exposure to low levels usually results in more severe and long-term symptoms than brief exposures to high levels.   This is because our bloodstream quickly mounts a defense against high levels of CO exposure, rapidly binding free CO to hemoglobin as carboxyhemoglobin [COHb].  Because COHb is a huge molecule, it effectively keeps CO from diffusing through capillaries into tissues where it could do much more damage.   The higher the level of CO exposure, the more quickly arterial COHb rises and the more quickly it reaches equilibrium with the [always lagging] venous COHb, after which no more CO is absorbed into tissues.   Studies done by Peterson and Stewart in the 1970s found that when breathing over 1000ppm, this equilibrium is reached in just a few hours.  But at very low levels of CO exposure, in the range of 10 to 25ppm, arterial and venous COHb rise very slowly and can take 20 hours or more to reach a-v  equilibrium.  And as long as arterial COHb remains greater than venous, the body is absorbing CO into tissues where it binds more aggressively than oxygen to other heme proteins such as myoglobin in muscle, neuroglobin in nerves, cytochromes in our liver, and all mitochondria. After the CO exposure ends, the CO that was thus absorbed into tissues takes much longer to dissociate and diffuse back out into venous blood.   A good rule of thumb is that it takes AT LEAST twice as long to exhale any given amount of CO as it took to inhale it. So from a toxicologist's perspective, any CO exposure that results in a large arterio-venous gap in COHb is extremely dangerous.   The a-v COHb gap in healthy people is usually very small, in the range of just 0 to 1%, but during CO exposures, post-mortem studies of CO victims show the a-v gap may be as high as 28%, and post-poisoning, the v-a gap [now going the other way] may be as high as 17%. To come back to the "danger" associated with specific CO levels, it is important to remember that exhaled CO levels in healthy people are usually UNDER 3ppm.   This CO comes not from exposures but from the routine breakdown of heme proteins by heme oxygenase into iron, biliverdin and CO.   When this endogenous CO production is increased--such as in people with diabetes or arthritis, or in women in the pre-menstrual phase of their period--the exhaled level is several times higher, in the range of 5-25pm.  I think we should consider any CO exposures that result in higher than normal levels of exhaled CO to be "dangerous" -- and deserving of daily treatment until the level of exhaled CO is back under 3ppm for a non-smoker. As for what level of inhaled CO is most dangerous, 100% pure CO kills mammals after just one breath.  This is why CO was commonly used by vets to euthanize animals in the 20th century, and by Nazis to kill Jews, Romas, homosexuals, and other "enemies of the state" in gas chambers before they switched to Zyklon B.