Probe traps are sensitive tools for detecting populations of beetles in stored grain, but their use for estimating insect density in concrete silos has not been established. Populations of grain beetles infesting wheat in concrete silos at two commercial storage facilities in north central Oklahoma were sampled using probe traps and grain probe samples over a 17-wk period. Grain temperature and quality parameters were collected during the study. Thirteen insect species were detected using probe traps, whereas eight species were detected with the grain samples; Cryptolestes ferrugineus and Rhyzopertha dominca were the most common insects collected in the grain samples. Across dates, there were no differences in mean number of insects recovered by species near the grain surface and 1 m below in either probe traps or grain samples. Comparison of slopes (insects recovered in grain samples regressed on insects in probe traps) showed that there were significant differences by trap position for C. ferrugineus but not R. dominica. Multiple regression models, developed to predict insect population estimates using number of insects captured in probe traps and a temperature component, indicated that more variability in the data were explained using traps positioned 1 m below the grain surface (R² = 0.70) than near the surface (R² = 0.21) for C. ferrugineus. About one-half (R² = 0.53) of the variability in insect density was described for R. dominca. These regression models show the potential for methodical use of grain probe traps in pest management decision-making.