Accurate estimates of cumulative NO fluxes from agricultural soil are essential for quantifying global NO emissions and for identifying effective mitigation strategies. This study focuses on the short term temporal variability of NO fluxes, and on how the accuracy of annual cumulative estimates is affected by different (simulated) measuring frequencies of an automatic system based on the closed chamber method. Fluxes were measured with high temporal resolution (24 per day) for approximately one year on agricultural soil in the southwest of Sweden. The short-term temporal variability of NO fluxes was considerable: the predictive power of measured fluxes decreased to negligible levels in a time frame of 4-5 h, and large intraday flux ranges were observed frequently, particularly in days with soil temperatures below C. A time of day well suited for consistently providing good estimates of the mean daily flux could not be identified. Consequently, the accuracy of annual emission estimates strongly depended on the measuring frequency. Multiple measurements per day were necessary for consistently quantifying annual emission estimates with accuracy. Accuracy improved with measuring frequency increasing up to four times per day, and higher frequencies generated negligible further improvements. Based on this study the following recommendations are made with regard to measuring frequency: i) measure four times per day for maximum accuracy and ii) measure twice per day for a good compromise between accuracy and the number of plots that can be monitored.