Health services research has become an important area for evaluating the cost effectiveness of interventions. When used in treatment outcome research, the accuracy of self-report data is essential. The reliability and validity of self-report service utilization among alcohol and drug addicted individuals is questionable and largely unexplored. This study assessed the accuracy of self-report utilization of services compared to service record abstraction in a sample of treatment seeking individuals with alcohol use disorders. The results of the comparative analysis found that the level of agreement for some services, particularly medical, psychiatric and substance abuse inpatient admissions, and social service involvement was good. There was less agreement in emergency room visits and arrests. Factors related to discrepancies between self-report and records were explored.