This is a critique of a study by Watson et al. on the validity of alcoholics' self-reports that suggested that clinical researchers and clinicians abandon the use of such data for now. It is shown, however, through an analysis of the rationale, procedures and data analyses reported by Watson et al., that their conclusions are not warranted. It is argued that the use of a situational and convergent validity approach to the collection and interpretation of self-report data is likely to continue to be a productive research and clinical strategy.