Evaluation of traffic noise is a typical acoustical task. Most design criteria are defined using metrics that are long-term (generally annual) averages; however, the daily, weekly, or seasonal variations in traffic noise are generally not reported. This variation should be included when calculating the uncertainty budget of measurements of traffic noise. Additionally, recent Green Building Design Standards in the United States, including school and healthcare facility design guidelines, require evaluation of traffic noise in terms of the loudest hourly L-eq. The authors have suggested a definition of the loudest hourly level that depends on the variance in the distribution of hourly levels (Proceedings of InterNoi se 2014). Therefore, for traffic noise measurements it is important not just to determine the average level but to characterize the distribution of levels of various time intervals. A measurement program was undertaken to obtain long-term traffic noise data with high time resolution. The authors analyze methods for estimating the variance of the distribution from short term data, and suggest minimum survey requirements for traffic noise measurements.