The role of reduced-impact logging on the preservation of biodiversity in tropical lowland dipterocarp rainforest is examined by looking at differences in dung beetle community structure between two sites logged in 1993: one harvested using reduced-impact methods and one using conventional techniques. Collections were made using two flight intercept traps over 7 d, and samples were compared with previous collections made from primary forest (riverine and interior-forest) and older forest logged using conventional techniques (logged in 1981). Of the two 1993 sites, the higher diversity and species richness (S = 57, n = 969, alpha = 13.23, H′ = 3.24) was recorded in the forest logged using reduced-impact techniques: the conventionally logged site had both lower diversity and species richness (S = 48, n = 1968, alpha = 8.88, H′ = 1.89), and lacked some primary forest specialists present in the reduced-impact forest samples. The dung beetle community in the 1993 conventional logged site is similar to that of the conventional logging site harvested in 1981. Primary forest has a well-defined ecotone, from interior to riverine forest: both 1993 sites contain a mixture of interior-forest and riverine specialists that are usually spatially separated along this ecotone. Although the dung beetle assemblage in the forest harvested by reduced-impact logging is more similar to a primary interior-forest assemblage than the conventionally logged site, both 1993 sites have assemblages that are closer in similarity to assemblages from primary-riverine habitat than they are to ones from primary interior-forest. However, because the forest logged using reduced-impact logging techniques has a more equitable and diverse dung beetle assemblage and a greater number of interior-forest specialists than the conventionally logged site, this study suggests that reduced-impact logging has better preserved the primary forest assemblage than conventional logging techniques.