This study develops an approach to measuring 'structural imbalance' in Australia's mental health sector, i.e. the non-correspondence between service use and mental illness diagnosis ('need'). The approach fully cross-classifies data on consumers, and non-consumers, of mental health services, with those having a diagnosis of mental illness, and those without a diagnosis. Venn diagrams are also constructed. The data are from Australia's epidemiological survey, Mental Health and Wellbeing... . 'People' are a proxy for resources (given that monetary data are unavailable). Extensive structural imbalance is found. Mentally ill people receiving no mental health services ('unmet need') is enumerated; likewise, people consuming mental health services who do not meet the criteria of mental illness ('met non-need'). Government-subsidised services are being used not just for mental illness but for performance (sport, executive etc), by 'the worried well', etc. Resource insufficiency in this sector is also found. Some policy options are discussed.