Background: Mental illness can impact all stages of labour market engagement: lower rates of participation in the labour market, higher rates of unemployment and employment in low-skill or low-earning occupations relative to qualifications. Systematic mental health surveys provide an opportunity to examine the scale of such impacts. Though usually cross sectional in nature, such surveys commonly include historical data by self report that can be used to construct a retrospective cohort study, within which it is possible to examine temporal sequence of illness and employment experience and thereby explore issues of causality., Methods: The 1997 National Survey of Mental Health and Wellbeing of Adults contains comprehensive questions relating to prevalence and level of disability associated with major mental disorders. Here we employ secondary analyses of the survey data to develop three models capturing workforce participation, unemployment and occupational level. Different versions of these models employ either broad diagnostic classes or numbers of disorders as indicators of mental disorder status. After reporting findings from these models we use them in combination to estimate labour market costs for Australia., Results: Each disorder reduces the chance of participation in the labour market by 1.3 percentage points, an appreciable amount given that most individuals suffering from mental disorders have multiple disorders. There is a strongly significant effect of mental illness on employment and clear evidence of reduced occupational skill level., Discussion: The impact of mental illness is very strong at every stage of engagement. Limitations include the self report nature of the assessments and lack of specific income data collection within the survey instrument., Implications for Health Care Provision and Use: Other work based on this survey shows poor accessibility of recovery based and rehabilitation orientated services. These are the very services that have a role to play in increasing workforce participation, employment and occupational level., Implications for Health Policies: This assessment of these labour market effects suggests that increasing mental health care funding could yield substantial benefits to the economy., Implications for Further Research: In 2007 a further survey of the Australian population was carried out that will provide an updated data set against which to re-examine this issue. The issue of comparability of the instrumentation between the two surveys will be complex, but valid comparisons across the two surveys should be feasible.