This study conducts a pilot test analyzing reports from New York, New Jersey, and California and uses textual analytics to reengineer government performance audit reporting. It advocates a performance audit database that can facilitate easier access and extract relevant information from lengthy reports in a timely manner. The study presents a framework to identify the commonalities and differences in terminologies used by sampled states, evaluates and extracts relevant content from the reports according to Generally Accepted Government Auditing Standards requirements, and constructs a taxonomy specific to government performance audits. Furthermore, this study investigates the disclosure quality by examining linguistic and similarity features, such as report length, specificity, readability, comprehensibility, and content similarity. This paper raises attention to a key legislative task that requires reporting reforms.