ABSTRACT Aims To examine (1) the structure of the Impaired Control Scale using confirmatory factor analytical techniques and (2) the internal consistency and validity of the ICS subscales. Design Cross-sectional data were collected on alcohol-related variables using standard self-completion scales. Participants Three hundred and fifty-seven social drinkers (not in alcohol treatment) and 137 treatment drinkers (in alcohol treatment) who acknowledged experiencing impaired control over alcohol intake in the last 6 months. Impaired control was defined as drinking more than intended or more than one thinks one should. Measurements Self-completion questionnaire consisting of demographic questions, Severity of Alcohol Dependence Data questionnaire (SADD), Impaired Control Scale (ICS) and three questions assessing alcohol consumption from the Alcohol Use Disorders Identification Test (AUDIT). Findings This first confirmatory factor analytical study of the ICS found that all three parts of the ICS were unidimensional for both social and treatment drinkers. These results confirmed previous exploratory principal components analysis findings. The excellent validity and internal consistency of all three parts of the ICS found here also confirmed results of previous research into the scale. Conclusions The confirmation of the structure and psychometric properties of the ICS suggest that the scale is a reliable and valid instrument, with a predictable and stable factor structure, for use with both clinical and non-clinical samples.