Because the microflora and arthropods in the leaf-litter community are sensitive to moisture, changes in rainfall predicted to occur with climate change (both increased rainfall and drought) will likely affect rates of litter decay. We constructed rainout shelters to cover eight fenced plots assigned to either a high- or low-rainfall treatment. We have manipulated rainfall 6-7 months/yr since June 2002. Litterbags of three mesh sizes, to exclude different faunal categories, were placed in each plot in autumn 2002. The rate of disappearance of leaves was 20% higher in the high-rainfall treatment during the first year. Arthropods are periodically sampled by sifting and litter extraction. Both techniques revealed that spiders have been 2x more abundant in high-rainfall than low-rainfall plots since the first post-manipulation census 5 months into the experiment. Litter extraction revealed that densities of Collembola, important microbivores, were 2x greater in high-rainfall plots after 5 months, but densities in the two treatments converged after 13 months. Larger Collembola, sampled by sifting litter, were 5x more abundant in the high-rainfall plots 17 months into the experiment. Litter extraction showed that mites were 2x more abundant in high-rainfall plots after 5 and 13 months. Despite clear effects of the rainfall perturbation on both rates of litter decay and the arthropod community, there was no interaction between rainfall and litterbag mesh size on rates of litter disappearance. Surprisingly, the rate of litter decay was similar in all three types of litterbag. The experiment is continuing for an additional year to determine if these patterns will persist.