Objectives While most UK military personnel transition successfully into civilian life, some experience unemployment and disability, which may be partly attributable to in-service factors. This study aims to determine the degree to which in-service mental health problems impact on postservice benefit claims.

Methods Using data from a cohort of 5598 recent leavers from regular service in the UK Armed Forces linked with data from the Department for Work and Pensions, we assessed associations between in-service mental health and postservice benefit claims, and the population attributable fraction (PAF) of benefit claims related to in-service mental health. An analysis with postservice mental ill health as mediator was performed to determine the degree to which the observed effects were a consequence of persistent illness, as opposed to remitted.

Results Mental illness occurring in-service predicted both unemployment and disability claims, partly mediated by postservice health (23%–52% total effects mediated), but alcohol misuse did not. Common mental disorder (CMD) (PAF 0.07, 95% CI: 0.02 to 0.11) and probable post-traumatic stress disorder (PTSD) (PAF 0.05, 95% CI 0.01 to 0.09) contributed to unemployment claims. Probable PTSD was the largest contributor to disability claims (PAF 0.25, 95% CI 0.13 to 0.36), with a smaller contribution from CMD (PAF 0.16, 95% CI 0.03 to 0.27).

Conclusions In-service mental ill health gives rise to benefit claims. These effects are only partly mediated by postservice mental health, implying that in-service (or pre-service) mental issues have carry-over effects into civilian life even if remitted. Better prevention and treatment of in-service PTSD symptoms may well reduce postservice disability claims.