Helicoverpa zea (Boddie) causes significant damage to several economically important crops throughout the United States. To control this insect and other lepidopteran pests, many host species have been genetically modified to express D-endotoxins from the soil bacterium Bacillus thuringiensis (Bt) Berliner. Populations of H. zea are potentially exposed to cry1Ac in transgenic cotton in the south and to cry1Ab in transgenic corn in the north. Widespread production of transgenic crops has raised concerns about targeted insects developing resistance to Bt and the potential for cross resistance. In Minnesota, an in-field screen using late-planted Bt sweet corn has been used since 1997 to detect insects with potential resistance to Bt and to monitor changes in the frequency of these individuals. Larvae ( 3^rd instar) actively feeding on Bt expressing tissues are of particular concern. Paired plots of transgenic sweet corn (event BT-11) and its non-Bt isoline (‘Bonus'), respectively, were planted at 5 locations throughout the southern half of the state. In 1997, no H. zea larvae ( 3^rd instar) were found after sampling 4,200 ears of Bt corn; nearby non-Bt sweet corn averaged 0.79 larvae/ear. In 1998, 131 late-instar larvae were recovered after sampling 19,744 ears; the average density of larvae on non-Bt sweet corn was 0.95 larvae/ear. In 1999, 33 late instars were found after sampling 36,120 ears; non-Bt sweet corn averaged 0.41 larvae/ear. Frequencies of late-instar larvae surviving on Bt sweet corn increased from 3.01x10^-4 in 1997 to 5.84x10^-3 in 1998 and declined to 2.3x10^-3 in 1999. As H. zea does not overwinter in Minnesota, different source populations (e.g., southern U.S.) with differing susceptibilities to Bt may account for variability in frequency estimates.