Elevated concentrations of carbon dioxide (CO2) prevented neonate larvae of the western corn rootworm, Diabrotica virgifera virgifera LeConte, from locating the roots of growing corn in behavioral bioassays conducted in soil tubs. When CO2 was pumped into one end of a soil tub, significantly more larvae were recovered from soil at the treated end than from soil around a growing corn plant at the opposite end of the tub. In controls with ambient air pumped into one end of a soil tub, significantly more larvae were recovered from the soil around the corn plant than from soil on the treated side. Larvae were unable to locate the roots of corn seedlings when CO2-generating materials were mixed into the soil. CO2-concentrations in soil were measured by mass spectrometry with selected ion monitoring at m/z 44. Granules composed of baker's yeast, yeast nutrients, and an organic substrate were prepared as a CO2 source and were tested in larger soil tub bioassays. Significantly fewer larvae were recovered from corn roots in the soil tubs with yeast granules than from corn roots in control soil tubs. The CO2-generating granules produced soil CO2 concentrations between 15.8 and 18.5 mmol/mol (compared with 1.7-2.6 mmol/mol in control tubs), and this was sufficient to prevent larvae from locating corn roots. In field trials, organic and inorganic CO2- generating treatments resulted in root ratings that were significantly lower than for the control plants.