AbstractBoreal lake sediments are carbon sources by producing CO2. CO2 flux from sediments is partly controlled by turbulence in the water column, which is not given the same attention as CO2 production rates in current estimates of CO2 fluxes from sediments. We quantified the in situ CO2 flux across the sediment‐water interface in a small (0.07 km2) lake in Sweden by measuring the in situ O2 flux with the Eddy Correlation (EC) method and using the apparent respiratory quotient (CO2 production:O2 consumption) derived from sediment incubations. We demonstrate that median CO2 flux estimated by EC was ~70% smaller than estimated by sediment incubations with artificial water mixing (1.0 × 10−2 and 3.6 × 10−2 µmol C m−2 s−1, respectively). Additionally, we show that inducing artificial mixing of supernatant water in the incubation experiment has a positive effect on observed fluxes, enhancing CO2 flux by ~30% compared to not mixing supernatant water. We suggest that the difference between the methods is due to the strong artificial water mixing in sediment incubations compared to the turbulent mixing in this small lake. Additionally, low O2 supply to sediment aerobic heterotrophic microbes during extended periods of low water currents can inhibit respiration and thus CO2 production. These findings suggest that the sediment contribution to total lake CO2 emission might currently be overestimated for small boreal lakes. Care should be taken when upscaling sediment CO2 flux derived from incubation experiments to entire basins of small lakes, as incubation experiments are unlikely to accurately mimic in situ bottom water currents and gas exchange.