Abstract: The evaluation method of capacity envelope curve based on historical data relied on the accuracy of sample data and prior information, resulting in poor robustness of evaluation results. In view of the shortcomings of this evaluation method, a new method of runway capacity evaluation based on ADS-B(Automatic Dependent Surveillance-Broadcast) data was proposed. By analyzing and mining the ADS-B data information that has been decoded and processed, the hourly arrivals and departures of the runway were statistically obtained. The K-S (Kolmogorov-Smirnov) test was applied to carry out hypothesis testing on the sample data obtained from ADS-B analytic data statistics. Box-Cox transformation converted non-normal data and interval estimation evaluated normal data. Taking Tianjin Binhai International Airport as an example, the capacity of arrival runway, departure runway and dual runway were calculated, and the proposed evaluation method was compared with the evaluation capacity based on envelope curve. It is found that the results of capacity evaluation based on envelope curve are 5% to 7.5% larger than the reference capacity value, while the new method of runway capacity evaluation based on ADS-B data can accurately cover the reference capacity value, and the evaluation results are more accurate and reliable. The results show that the proposed new capacity evaluation method can effectively reduce the impact of sample data differences on the evaluation results and has better robustness. Compared with the reference capacity, the new runway capacity evaluation method based on ADS-B data can provide a reference for air traffic controller to further optimize runway capacity.