With the recent advent of Intelligent Transportation Systems (ITS), and their associated data collection and archiving capabilities, there is now a rich data source for transportation professionals to develop capacity values for their own jurisdictions. Unfortunately, there is no consensus on the best approach for estimating capacity from ITS data. The motivation of this paper is to compare and contrast four of the most popular capacity estimation techniques in terms of (1) data requirements, (2) modeling effort required, (3) estimated parameter values, (4) theoretical background, and (5) statistical differences across time and over geographically dispersed locations. Specifically, the first method is the maximum observed value, the second is a standard fundamental diagram curve fitting approach using the popular Van Aerde model, the third method uses the breakdown identification approach, and the fourth method is the survival probability based on product limit method. These four approaches were tested on two test beds: one is located in San Diego, California, U.S., and has data from 112 work days; the other is located in Shanghai, China, and consists of 81 work days. It was found that, irrespective of the estimation methodology and the definition of capacity, the estimated capacity can vary considerably over time. The second finding was that, as expected, the different approaches yielded different capacity results. These estimated capacities varied by as much as 26 % at the San Diego test site and by 34 % at the Shanghai test site. It was also found that each of the methodologies has advantages and disadvantages, and the best method will be the function of the available data, the application, and the goals of the modeler. Consequently, it is critical for users of automatic capacity estimation techniques, which utilize ITS data, to understand the underlying assumptions of each of the different approaches.