Freeze-thaw fluctuations in soil temperature may be critical events in the annual pattern of nutrient mobilisation that supplies plant growth requirements in some temperate, and most high latitude and high altitude ecosystems. We investigated the effects of two differing freeze-thaw regimes, each of which is realistic of in situ spatial and temporal variation in field conditions, on C and N dynamics in sub-arctic heath tundra mesocosms. In addition, N-15 isotopic label was used to follow the partitioning of a labile N pool between major ecosystem components, both during the freeze-thaw treatments phase, and in a subsequent equilibration phase. A single deep freeze treatment phase enhanced dissolved total and labelled N pools in the soil solution at initial thaw, and resulted in reduced pool sizes at the end of the equilibration phase. By contrast, a multiple freeze-thaw cycling treatment directly enhanced the dissolved labelled N pool, but did not significantly affect dissolved total N. Furthermore, both dissolved labelled N and dissolved total N pools were significantly enhanced in the equilibration period following multiple freeze-thaw, the latter due to a marked increase in soil solution NH4+. Microbial biomass C was not significantly affected by either of the freezing treatments upon final thaw, but was significantly reduced over the combined treatment and equilibration phases of the multiple freeze-thaw regimes. Furthermore, the treatments had no significant effects on total or labelled N within the microbial biomass over either phase. Total mesocosm CO2 efflux rates remained closely correlated with soil temperature throughout the experiment in both regimes, suggesting that respiratory flushes associated with treatment-induced microbial cell lysis were negligible. Together, these results indicate that moderate freeze-thaw fluctuations may have minimal influences on microbial biomass pools, but nevertheless can have strong contrasting effects on the amounts, forms, and timing of N and organic C supply into the soil solution. Ecosystem losses via N2O effluxes were of greatest magnitude immediately upon thawing in both treatments, and were of similar total magnitude to inorganic N leachates in throughflow. Herb leaves, total fine roots, and vascular stems accumulated some N-15 label in one or both of the freezing treatments by the end of the experiment. Together, these results indicating very small N losses relative to the magnitudes of internal transfers, suggest tight ecosystem N cycling both during and after freeze-thaw events. Furthermore, our small and subtle effects on microbial and soluble C and N pools relative to previous studies using more severe regimes, suggests that periods of moderate freeze-thaw fluctuations may have only a minor influence on the annual pattern of C and nutrient dynamics in seasonally cold ecosystems. (C) 2004 Elsevier Ltd. All rights reserved.