The purpose of this study was to assess the persistence and accumulation of the Cry1Ab protein in soil as a result of sustained planting of genetically modified Bacillus thuringiensis (Bt) corn hybrids. Soil samples were collected from agricultural fields in five corn-growing regions of the United States where Bt corn hybrids (MON 810 or Bt11) had been planted for at least 3 consecutive yr. At each site, soil samples were collected during the corn-growing period (postanthesis) and again within 6 wk after harvest. Multiple soil specimens from matched Bt cornfields and nearby, non-Bt control fields were analyzed by diet-incorporation insect bioassay, using growth inhibition (GI) of the European corn borer (Ostrinia nubilalis) as the toxicity endpoint. Positive control soil samples containing Cry1Ab protein at the GI₅₀ level (0.05 μg/g soil) were analyzed in tandem with test and control samples to verify that the bioassay was able to detect low levels of Cry1Ab protein. The limit of detection for Cry1Ab protein in soil was 0.03 μg/g soil. The presence of Cry1Ab protein in soil was assessed by statistical comparison of the insect toxicity (GI) of soils collected from Bt and non-Bt (control) cornfields. Only one soil sample, collected postanthesis in a Bt cornfield that had also been treated with carbofuran insecticide, showed insect toxicity. This toxicity was below the GI₅₀ level, and no toxicity was detected in the soil collected from the same plot shortly after harvest. Therefore, there is no evidence of persistence or accumulation of Cry1Ab protein in soils from fields planted for at least three consecutive growing seasons with Bt corn hybrids.