Abstract. Experimental efforts to determine how insects influence terrestrial wood decomposition are few, especially in temperate regions. To address this need, a five-year exclusion study was conducted in northern Mississippi, U.S.A., to quantify insect contributions to wood decay using one-meter loblolly pine (Pinus taeda L.) bolts. The study included three treatments: (1) ''partially protected'' bolts that were placed on cypermethrin-treated soil to exclude subterranean termites (Isoptera: Rhinotermitidae: Reticulitermes spp.) while permitting colonization by beetles (Coleoptera) and other saproxylic taxa, (2) ''fully protected'' bolts that were placed on cypermethrin-treated soil and enclosed within screen cages to protect against all insects and (3) ''unprotected'' bolts that were not subjected to either exclusion treatment. The full insect community consumed approximately 15-20% of wood volume in unprotected bolts, about six times more than in partially protected bolts from which termites were excluded. There were no differences in specific gravity (based on initial wood volume) or mass loss among treatments, however. It is not clear whether these findings are due to an inhibition of microbial decomposers by insects (e.g., antimicrobial compounds secreted by termites or ants), a stimulatory effect of the exclusion treatments (e.g., cypermethrin stimulating fungal growth or cages favorably altering wood moisture), or some combination of both. When based on final water-displaced volume, specific gravity was significantly higher for unprotected bolts than for those fully protected, probably because termites selectively consume the least dense wood. By the end of the study, about 20% of the final dry weight of unprotected bolts consisted of termite-imported soil. Wood volume consumed and soil content decreased with distance from the ends of the bolts whereas water content exhibited the opposite pattern. We detected a significant negative relationship between water content and volume consumed by termites, possibly because water content decreases with increasing wood density and termites tend to avoid high density wood. While insects clearly consume large volumes of wood in southeastern U.S. forests, our results suggest they do not act to accelerate mass loss beyond what is achieved by microbial decomposers. More research is needed to confirm this, however-especially given the uncertainties inherent to exclusion studies.