Relaxed eddy accumulation (REA) systems that employ one single long inlet tube are prone to measurement uncertainties caused by (a) an imprecisely determined lag time between the change of sign in the vertical wind velocity and the switching of the splitter valves and (b) attenuation of high-frequent concentration fluctuations in the tube flow. However, there is currently no commonly applied procedure to address these uncertainties. In this study, we first evaluated the lag time error of the volume flow, mass flow and cross-correlation method (online and offline) and experimentally determined the magnitude of high frequency attenuation for a 21.5 m long inlet tube of an operating REA system. In a second step, we simulated the impact for different artificial lag time errors and low-pass filter strengths on the REA concentration differences and, thus, on the REA flux, using high frequency time series of temperature, O₃, CO₂ and H₂O. The reduction of scalar fluxes was mainly correlated with increasing switching frequencies and ranged for typical lag time errors of the investigated REA system between <5% and 50%, whereas the flux loss due to high frequency attenuation was between <5% and 30%. The results were very similar for all scalar quantities. Based on our results, we derived empirical correction functions for both imprecise lag times and high frequency attenuation, discuss their potential application to correct fluxes measured with other REA systems and give a general procedure to address the uncertainties in future REA setups.