Due to concerns about data quality, Automated Passenger Counting technology has rarely been used to analyze local ridership trends. This paper presents a novel framework to test the consistency and completeness of passenger count data in four cities. The data are aggregated at the system level and compared with the National Transit Database between 2012 and 2018. In all four agencies, passenger counts closely follow the fluctuations observed in the national transit database. There is, however, a slight drift in two of the four agencies. At the stop-level, missing and duplicate vehicle-trips are identified using schedule data from the General Transit Feed Specification. Missing and duplicate trips only concern a small proportion of stops, which can be eliminated using the proposed method. Overall, this research leads the way towards the analysis of factors affecting ridership on a tight spatial and temporal scale.