The ecosystem-scale exchange fluxes of energy, water and carbon dioxide (CO2) between wet arctic tundra and the atmosphere were investigated by the micrometeoro-logical eddy covariance method. The investigation site was situated in the centre of the Lena River Delta in Northern Siberia (72°22'N, 126°30'E). The micrometeorological campaigns were performed from July to October 2003 and from May to July 2004. The study region is characterised by a polar and distinctly continental climate, very cold and ice-rich permafrost and its position at the interface between the Eurasian continent and the Arctic Ocean. The measurements were performed on the surface of a Holocene river terrace, which is characterised by wet polygonal tundra. The soils at the site are characterised by high organic matter content, low nutrient availability and pronounced water logging. The vegetation is dominated by sedges and mosses. The fluctuations of the wind velocity components and the sonic temperature were determined with a three-dimensional sonic anemometer, and the fluctuations of the H2O and CO2 concentrations were measured with a closed-path infrared gas analyser. The measurement height was 3.65 m. The fast-response eddy covariance measurements were supplemented by a set of slow-response meteorological and soil-meteorological measurements. The relative energy balance closure was around 90 % on the hourly basis and around 96 % on the daily basis, indicating a good performance of the complete flux measurement set-up. The combined datasets of the two campaigns 2003 and 2004 were used to characterise the seasonal course of the energy, water and CO2 fluxes and the underlying processes for the synthetic measurement period May 28…October 21 2004/2003 which included the period of snow and soil thawing as well as the beginning of refreezing. The synthetic measurement period 2004/2003 was characterised by a long snow ablation period (until June 17) and a late start of the growing season. On the other hand, the growing season ended also late due to high temperatures and snow-free conditions in September. The cumulative summer (June…August) energy partitioning was characterised by low net radiation (607 MJ m-2), large ground heat flux (163 MJ m-2), low latent heat flux (250 MJ m-2) and very low sensible heat flux (157 MJ m-2) compared to other tundra sites. These findings point out the major importance of the very cold permafrost (due to extreme winter cooling) for the summer energy budget of the tundra in Northern Siberia. The partitioning of the available energy into latent and sensible heat fluxes was typical for arctic wetlands as indicated by the Bowen ratio, which ranged between 0.5 and 1.5 during most of the summer. Despite a high cumulative precipitation of 201 mm during summer (June…August), the cumulative summer evapotranspiration of 98 mm (mean (1.1 ± 0.7) mm d-1) was low compared to other tundra sites. The water exchange between the arctic wetland and the atmosphere was normally limited by the low available energy and only seldom constrained by low water availability. The average decoupling factor Ω of 0.53 ± 0.13 indicated a relatively low coupling of the atmosphere and the vegetation compared to other tundra ecosystems. In summer 2003, heavy rainfall initiated severe thermoerosion phenomena and in the consequence increased drainage and run-off at the wet polygonal tundra thus demonstrating the sensitivity of permafrost landscapes to degradation by changes in hydrology. The CO2 budget of the wet polygonal tundra was characterised by a low intensity of the main CO2 exchange processes, namely the gross photosynthesis and the ecosystem respiration. The gross photosynthesis accumulated to -432 g m-2 over the photosynthetically active period (June…September). The contribution of mosses to the gross photosynthesis was estimated to be about 40 %. The diurnal trend of the gross photosynthesis was mainly controlled by the incoming photosynthetically active radiation (PAR) with the functional response well described as a rectangular hyperbola. During midday the photosynthetic apparatus of the canopy was frequently near saturation and represented then the limiting factor on gross photosynthesis. The seasonal progression of the gross photosynthesis was controlled by the combination of the phenological development of the vegetation and the general temperature progression over the summer. Water availability was only of minor importance as control on the gross photosynthesis due to the wet soil conditions at polygonal tundra. However, the gross photosynthesis was temporarily significantly reduced when the mosses at the drier microsites of the polygon rim experienced water stress during longer periods of advection of warm and dry air from the South. The synoptic weather conditions affected strongly the exchange fluxes of energy, water and CO2 by changes in cloudiness, precipitation and the advection of air masses from either the Siberian hinterland or the Arctic Ocean. The ecosystem respiration accumulated to +327 g m-2 over the photosynthetically active period, which corresponds to 76 % of the magnitude of the gross photosynthesis. However, the ecosystem respiration continued at substantial rates during autumn when photosynthesis had ceased and the soils were still largely unfrozen. The temporal variability of the ecosystem respiration during summer was best explained by an exponential function with surface temperature, and not soil temperature, as the independent variable. This was explained by the major role of the plant respiration within the CO2 balance of the tundra ecosystem. The wet polygonal tundra of the Lena River Delta was observed to be a substantial CO2 sink with an accumulated net ecosystem CO2 exchange of -119 g m-2 over the summer and an estimated annual net ecosystem CO2 exchange of -71 g m-2. The analysis of the qualitative relationships between the processes and environmental factors, which control the energy, water and CO2 budget, suggested that the wet arctic tundra will experience severe perturbations in response to the predicted climatic change. The alterations of the tundra ecosystems would in turn exert pronounced mainly positive feedbacks on the changing climate on the regional and global scale.