The paper investigates the sensitivity of sonic anemometer-derived stress estimates to the tilt of the anemometer, particularly in unstable stratification and deep convective boundary layers. Three methods for determining tilt angles and computing tilt-corrected stresses are compared. The commonly used double rotation method is shown to introduce significant run-to-run stress errors due to sampling uncertainty in the mean vertical velocity. The triple rotation method, while more accurate, still results in large errors due to combined sampling errors. For sea measurements where cross-stream stress is significant, the double rotation method overestimates surface stress due to uncorrected lateral tilt. A third method, using a planar fit technique, reduces run-to-run stress errors and provides an unbiased estimate of lateral stress. The paper emphasizes the importance of using a mean streamline coordinate system for analyzing surface-layer turbulence data over sloping terrain to minimize the impact of terrain slope on stress measurements.The paper investigates the sensitivity of sonic anemometer-derived stress estimates to the tilt of the anemometer, particularly in unstable stratification and deep convective boundary layers. Three methods for determining tilt angles and computing tilt-corrected stresses are compared. The commonly used double rotation method is shown to introduce significant run-to-run stress errors due to sampling uncertainty in the mean vertical velocity. The triple rotation method, while more accurate, still results in large errors due to combined sampling errors. For sea measurements where cross-stream stress is significant, the double rotation method overestimates surface stress due to uncorrected lateral tilt. A third method, using a planar fit technique, reduces run-to-run stress errors and provides an unbiased estimate of lateral stress. The paper emphasizes the importance of using a mean streamline coordinate system for analyzing surface-layer turbulence data over sloping terrain to minimize the impact of terrain slope on stress measurements.