Data from the Australian Longitudinal Survey, conducted by the Commonwealth Department of Employment, Education and Training, were analysed to estimate relative risk of psychological disturbance accompanying unemployment in young people aged 15-24 years. Two cohorts were surveyed annually over 4 years during the mid-to-late-1980s; one from the general 15-24 year-old population (N = 8995), and the other selected from Commonwealth Employment Service records (N = 2403). Such large respondent numbers allowed control of confounding by exclusion to isolate employment transition groups suitable for hypothesis testing and quantification of causal relationships. Psychological morbidity was measured using binary outcomes of the 12 item psychological component of the General Health Questionnaire. Excluded from the analysis were those who: suffered from pre-existing physical health problems; were dissatisfied in their job; were self-employed; underwent marriage breakdown during the inter-survey period; had become widowed during the inter-survey period. A Bayesian probabilistic approach was used to calculate probabilities of psychologically normal respondents becoming psychologically morbid, given prior transition from employment to unemployment. Mantel-Haenszel analysis was utilised to estimate relative risks in comparison to a control group of those remaining employed, after controlling for age and gender. An overall relative risk of becoming psychologically disturbed as a consequence of becoming unemployed was estimated to be 1.51 (95% CI: 1.15-1.99). The overall relative risk of recovery from psychological disturbance upon re-employment in those with psychological disturbance was estimated to be 1.63 (95% CI: 1.08-2.48). Residual psychological effects of past unemployment experience and the effects of long-term unemployment were investigated, but found to be non-significant in this study. There was some evidence of psychological adaptation to unemployment, but this was statistically insignificant. Unemployment was a significant cause of psychological disturbance in young people who were initially employed, not suffering physical ill-health, and psychologically normal; conversely, re-employment reversed the effect.