CO is a leading cause of poisoning deaths in the US today. Treating physicians use the carboxyhemoglobin (COHb) % saturation to guide the diagnosis and treatment of CO intoxication. We conducted a telephone survey of hospitals and laboratories in the Boston area, focusing on methodology for COHb determination and accompanying COHb reference intervals. Among 130 facilities, 23 (18%) provide COHb analysis. All facilities that perform the COHb test utilize dedicated multiwave-length photometry. Reference intervals for COHb varied widely among facilities. Eight of 21 (38%) facilities give unacceptably high "normal intervals" for nonsmokers when compared with values available in the literature. Thirteen of 20 (65%) use reference intervals for smokers that are too low, and 3 of 20 (15%) use values that are too high. These reference values provided by the testing facilities may be misleading to the ordering physicians unfamiliar with background COHb saturations. This may lead to misdiagnoses, false reassurances, and perhaps less aggressive treatment than might be warranted. The results of this study argue for wider adoption of COHb reference intervals supported by the current literature.