Abstract. Fluxes of greenhouse gases (GHG) carbon dioxide (CO 2 ), methane (CH 4 ) and nitrous oxide (N 2 O) were measured during a two month campaign at a drained peatland forest in Finland by the eddy covariance (EC) technique (CO 2 and N 2 O), and automatic and manual chambers (CO 2 , CH 4 and N 2 O). In addition, GHG concentrations and soil parameters (mineral nitrogen, temperature, moisture content) in the peat profile were measured. The aim of the measurement campaign was to quantify the GHG fluxes during freezing and thawing of the top-soil, a time period with potentially high GHG fluxes, and to compare different flux measurement methods. The forest was a net CO 2 sink during the two months and the fluxes of CO 2 dominated the GHG exchange. The peat soil was a small sink of atmospheric CH 4 and a small source of N 2 O. Both CH 4 oxidation and N 2 O production took place in the top-soil whereas CH 4 was produced in the deeper layers of the peat, which were unfrozen throughout the measurement period. During the frost-thaw events of the litter layer distinct peaks in CO 2 and N 2 O emissions were observed. The CO 2 peak followed tightly the increase in soil temperature, whereas the N 2 O peak occurred with a delay after the thawing of the litter layer. CH 4 fluxesCorrespondence to: M. K. Pihlatie (mari.pihlatie@helsinki.fi) did not respond to the thawing of the peat soil. The CO 2 and N 2 O emission peaks were not captured by the manual chambers and hence we conclude that high time-resolution measurements with automatic chambers or EC are necessary to quantify fluxes during peak emission periods. Subcanopy EC measurements and chamber-based fluxes of CO 2 and N 2 O were comparable, although the fluxes of N 2 O measured by EC were close to the detection limit of the system. We conclude that if fluxes are high enough, i.e. greater than 5-10 µg N m −2 h −1 , the EC method is a good alternative to measure N 2 O and CO 2 fluxes at ecosystem scale, thereby minimizing problems with chamber enclosures and spatial representativeness of the measurements.