Freeze-thaw (F-T) processes are prevalent and have the inherent potential to alter soil respiration (SR) in cold regions, which are particularly sensitive to climate warming and contribute a large uncertainty to the global carbon budget. However, the impacts of F-T processes on the diel pattern and temperature hysteresis of SR remain unclear. In this study, the Fick's law-based gradient method was employed to obtain high-temporalresolution SR data, and a four-parameter Gaussian function was used to quantify the diel pattern of SR. The results demonstrate that the magnitude and timing of the daily SR peak exhibit considerable variability across different F-T stages, leading to a delay in the optimal time window for measuring the daily mean SR during the freezing and frozen stages. Furthermore, the daily peak value of SR is primarily regulated by soil temperature (ST) during the thawed period, whereas soil water content (SWC) exerts a more pronounced effect than ST throughout the entire freezing-frozen-thawing period. The diel hysteresis between SR and ST was observed yearround, but its direction was reversed by F-T processes. Additionally, the magnitude of the temperature hysteresis was negatively correlated with SWC. These findings have important implications for understanding the mechanisms driving SR variability and improving the accuracy of annual carbon budget estimates in cold regions.