Pitfall traps were used to sample beetles (Coleoptera) in plots with or without inputs of dead loblolly pine ( Pinus taeda L.) wood at four locations (Louisiana, Mississippi, North Carolina and Texas) on the coastal plain of the southeastern United States. The plots were established in 1998 and sampling took place in 1998, 1999, and 2002 (only 1998 for North Carolina). Overall, beetles were more species rich, abundant and diverse in dead wood addition plots than in reference plots. While these differences were greatest in 1998 and lessened thereafter, they were not found to be significant in 1998 due largely to interactions between location and treatment. Specifically, the results from North Carolina were inconsistent with those from the other three locations. When these data were excluded from the analyses, the differences in overall beetle richness for 1998 became statistically significant. Beetle diversity was significantly higher in the dead wood plots in 1999 but by 2002 there were no differences between dead wood added and control plots. The positive influence of dead wood additions on the beetle community can be largely attributed to the saproxylic fauna (species dependent on dead wood), which, when analyzed separately, were significantly more species rich and diverse in dead wood plots in 1998 and 1999. Ground beetles (Carabidae) and other species, by contrast, were not significantly affected. These results suggest manipulations of dead wood in pine forests have variable effects on beetles according to life history characteristics.