A study was conducted in 1986 and 1987 to determine the effects of perturbation and habitat characteristics on the population dynamics and community structure of darkling beetles on mixed-grass rangeland. Two 1,400-ha blocks of rangeland in western South Dakota were treated aerially with malathion liquid spray or carbaryl-bran bait in early July 1986 to determine the immediate and second-year effect of treatments on darkling beetles, as well as grasshoppers and other nontarget arthropods. Populations of the dominant species (83% of total),Eleodes opacus(Say), were reduced significantly by 59.2 and 87.9% in the carbaryl-bran bait and malathion spray treatment plots, respectively, within 1 wk of treatment and remained at low levels throughout the rest of the summer. Populations did not change significantly in the control plots. Populations of the next three most abundant species (15.7% of total),E. tricostatus(Say),E. obsoleta(Say), andE. suturalis(Say), were negatively affected by one or both treatments. Populations increased to or above pretreatment levels in 1987 in treated and control plots. Analysis of covariance of the abundance of darkling beetles during July 1987, with pretreatment abundance as a covariable, indicated that treatment (treatment and block effects combined) had no significant effect on second-year populations ofE. opacus, E. obsoleta, orE. suturalis, but did explain a significant amount of variation of second-year populations ofE. tricostatus. However, the covariable, pretreatment abundance, was significantly correlated with July 1987 population levels ofE. opacus, E. tricostatus, andE. obsoleta. The multivariate technique, detrended correspondence analysis (DCA), was used to ordinate undisturbed communities along ecological gradients, and correlation analysis of DCA axes values with habitat variables was used to define specific gradients underlying the community structure of darkling beetles. Analysis of abundance and incidence data indicated that darkling beetle communities were ordinated along ecological gradients of percentage of grass coverage, particularlyBuchloe dactyloides(Nutt.) Engelm. andBromusspp., coverage of several forb species, and percentage of sand and clay in the soil. Numbers of total darkling beetles,E. opacus, andE. obsoleta, and species richness increased along a gradient defined by coverage of grass and percentage of sand.E. suturalisdecreased along the same gradient.