Subterranean termites, Reticulitermes spp., were attracted to carbon dioxide (CO2) in laboratory and field tests. In behavioral bioassays, Reticulitermes flavipes (Kollar), Reticulitermes tibialis Banks, and Reticulitermes virginicus Banks were attracted to CO2 concentrations between 5 and 50 mmol/mol. In further bioassays, R. tibialis and R. virginicus were attracted to the headspace from polyisocyanurate construction foam that contained 10-12 mmol/mol CO2. In soil bioassays in the laboratory, more termites foraged in chambers containing CO2-generating formulations than in unbaited control chambers. In field tests, stations containing CO2-generating baits attracted R. tibialis away from wooden fence posts at rangeland sites in Colorado. For all of the CO2 formulations tested, termites foraged in significantly more bait stations at treatment fenceposts than in bait stations at the control fenceposts. By the end of the 8-wk study, the number of bait stations located by termites at treatment fenceposts ranged from 40 to 90%. At control fenceposts, termites foraged in only a single station and the one positive station was not located by termites until week 5 of the study. At treatment fenceposts, termites foraged equally in active stations (containing a CO2-generating bait) and passive stations (with no CO2-generating bait), indicating that bait stations may benefit passively from a proximal CO2 source in the soil. CO2 used as an attractant in current baiting systems could improve their effectiveness by allowing earlier exposure of termites to an insecticide.