The effect of temperature changes on soil communities is an important aspect when estimating the effects of a predicted climate change. The aim of this investigation was to increase knowledge on how freeze-thaw cycles alter the soil microarthropod community in the sub-arctic. The abundance of springtails and mites was investigated after three seasons of prolonged periods of freeze-thaw cycles in the field, and the presence or absence of migration barriers, at two different field sites. Dome shaped transparent plastic greenhouses were successfully used as a novel method to increase freeze-thaw cycle frequencies in the soil. At a fellfield site, freeze-thaw treatment did not lead to significant differences in the five main soil faunal groups, but increased abundance were seen in a number of separate taxa. There was no freeze-thaw treatment effect on soil microbial biomass or soil nutrients, although treatments interacted as inorganic N increased in the separate freeze-thaw and migration barrier treatments. By contrast, at a glade site responses were strong due to more pronounced increases in the number of freeze-thaw cycles. The highest numbers of Collembola after 2 years of treatment were found in the freeze-thaw plots, in combination with migration barriers. The freeze-thaw treatment here also resulted in more Oribatida, microbial biomass C and dissolved organic C. A common hypothesis is that an increased number of freeze-thaw cycles would result in elevated winter mortality in microarthropods due to increased risk of inoculative freezing. However, we observed no increased mortality due to freeze-thaw events. Rather, there was a stimulation of soil microarthropods and microbial biomass, perhaps due to a prolonged period of microbial and faunal activity when the soil is repeatedly frozen and thawed compared to a constantly frozen soil. (c) 2007 Elsevier B.V. All rights reserved.