AbstractCurrent and future warming of high‐latitude ecosystems will play an important role in climate change through feedbacks to the global carbon cycle. This study compares 6 years of CO2 flux measurements in moist acidic tundra using autochambers and eddy covariance (Tower) approaches. We found that the tundra was an annual source of CO2 to the atmosphere as indicated by net ecosystem exchange using both methods with a combined mean of 105 ± 17 g CO2 C m−2 y−1 across methods and years (Tower 87 ± 17 and Autochamber 123 ± 14). The difference between methods was largest early in the observation period, with Autochambers indicated a greater CO2 source to the atmosphere. This discrepancy diminished through time, and in the final year the Autochambers measured a greater sink strength than tower. Active layer thickness was a significant driver of net ecosystem carbon exchange, gross ecosystem primary productivity, and Reco and could account for differences between Autochamber and Tower. The stronger source initially attributed lower summer season gross primary production (GPP) during the first 3 years, coupled with lower ecosystem respiration (Reco) during the first year. The combined suppression of GPP and Reco in the first year of Autochamber measurements could be the result of the experimental setup. Root damage associated with Autochamber soil collar installation may have lowered the plant community's capacity to fix C, but recovered within 3 years. While this ecosystem was a consistent CO2 sink during the summer, CO2 emissions during the nonsummer months offset summer CO2 uptake each year.