Background: A 12-month study of UK supported employment providers found that 77 (54.6%) of the participants in the study remained unemployed, 32 (22.7%) got jobs and 32 (22.7%) retained the jobs they held at the outset. Aims: To explore the impact of moving into employment on service use, earnings, benefits and tax allowances claimed. Methods: Service use and frequency were measured at baseline and 12 months. Comparisons paid particular attention to the differences between people entering work and those who remained unemployed. Costs were analysed from a government perspective (excluding earnings) and a societal perspective (excluding welfare benefits and taxes). Results: People who entered work reduced their consumption of mental health services (p<0.001). However, use of supported employment increased (p=0.04), in contrast to falling use by people who remained unemployed (p<0.001) and those who had been working for more than one year (p=0.002). The increase in earnings for those entering work (p=0.02) was not offset by a similar reduction in benefits. Conclusion: This indicates that mental health services may make savings as a result of their clients engaging in paid work. It raises questions about the optimal nature and organization of employment support for this service user group.