Abstract Insect populations infesting wheat stored in four bins on two Kansas farms were monitored from early July 1996 through to the middle of January 1997. Estimates of adult insect density based upon the numbers of adult insects caught using probe traps differed from those based upon the number of insects found in grain samples. These differences were a result of differences in numbers of insects found and percentages of traps or grain samples with insects. Traps detected insects 15 to 37 d earlier than grain samples. The depth of traps below the grain surface tended to influence both the total number and species composition of the insects that were caught. Traps inserted with the top just below the grain surface collected an average of 1.9 times more Cryptolestes ferrugineus (Stephens), 1.2 times more Rhyzopertha dominica (F.), 4.1 times more Ahasverus advena (Waltl) and 77.4 times more Typhaea stercorea (L.) than traps inserted with the top 7.6 cm below the grain surface. However, trap depth did not have a significant effect on the number of R. dominica caught and on only 12 to 21% of sampling dates did trap depth have a significant effect on the number of insects of other species that were caught. Grain temperatures in three of the bins averaged 30 °C during the first 70 d of storage and then decreased by 0.2°C/d. Grain in the other bin was initially more than 10°C warmer and grain temperature decreased by 0.2°C/d over the full storage period. The numbers of insects captured in traps decreased as grain temperature decreased even though grain samples indicated that insect populations were still growing. Thus, trap catches did not estimate insect population density consistently throughout the storage period. A method was developed in the current paper to adjust for the effect of seasonal changes in temperature on trap catch.