Ecosystem respiration (R_eco) plays an important role in estimating the global carbon balance of terrestrial ecosystems, and knowledge of respiration is required to correctly assess the gross primary (GPP) production of such ecosystems. In our paper, daytime R_eco fluxes are compared with nighttime CO₂ fluxes which were measured manually by the closed chamber technique on a wetland site in Poland. The CO₂ fluxes measurements were performed between July 2008 and May 2009. Our study indicated that there are significant differences between modeled daytime and nighttime R_eco fluxes and that these differences are higher when the nighttime fluxes are restricted only to nights with stable atmospheric conditions. The nighttime sums of cumulative R_eco are much higher than the one estimated for daytime conditions, while there is nearly no difference between daytime and nighttime sums of cumulative R_eco, when the nighttime fluxes are restricted only for nights with turbulent atmospheric conditions. Consequently, R_eco models developed on the basis of nighttime data series can overestimate CO₂ fluxes when they are used to estimate daytime respiration. In order to reduce uncertainties in estimation of daily R_eco fluxes on the basis of the nighttime chamber measurements, the nighttime R_eco models should be restriced only to the fluxes that were measured in turbulent atmospheric conditions. The biases in nighttime chamber measurements (especially when conducted in stable atmospheric conditions) as well as differences in estimated daytime and nighttime R_eco fluxes can have a significant effect on the assessment of global carbon balances of terrestrial ecosystems.