Background Academic literature and government initiatives have emphasised the importance of work as a means of improving health and reducing reoffending among offenders with mental disorders. Whilst a number of work skills programmes have shown promise for offenders more generally, evaluation of evidence for their effectiveness for those with a mental disorder is lacking, particularly in relation to improving employment outcomes. Aims To assess the evidence on the effectiveness of work skills programmes for mentally disordered offenders. Method A systematic review of the literature was conducted by searching the following databases: PsycINFO, CINAHL, Cochrane Library (Trials Register), Embase and Medline, using search terms which included Work Skills Programme*, Offend* and Mental*. Any empirical comparison study of work skills programmes was included in this review. The primary outcome was employment. Secondary outcomes included employment outcomes, reoffending, education, mental state, substance misuse, global functioning, quality of life, acceptability, leaving the study early and cost effectiveness or other economic outcomes. Results Six articles met the inclusion criteria. Collectively they provided limited evidence that work skills programmes increase the likelihood of people with mental disorder who are offenders obtaining employment in the short term, but there are insufficient studies to determine the long‐term impact of work skills programmes. Conclusions There is modest evidence to support inclusion of specific work skills programmes in the treatment of offenders with mental disorder. Future studies should be of theoretically driven programmes, such as Individual Placement Support (IPS), and use a standard set of relevant outcome measures and long enough follow‐up for testing the effectiveness of any programme on engagement in competitive, paid employment as, even if skilled, offenders with mental disorder must constitute a hard to place group. Copyright © 2015 John Wiley & Sons, Ltd.