Research consistently reports a veteran mortality disadvantage relative to nonveterans, but has not considered the contribution of service-connected disability to this differential.

This study uses data from the 1986 and 1989 National Health Interview Survey-2011 Linked Mortality Files (N = 124,122) to estimate multivariate Cox regression models of the association between veteran status and mortality, taking service-connected disability status into account. Bivariate analyses demonstrate higher mortality risk, lower socioeconomic status, and poorer health and functioning among veterans with a service-connected disability than among nonveterans and veterans without a service-connected disability. Multivariate models confirm a mortality disadvantage for all veteran service-connected disability subgroups, which is reduced by the inclusion of exogenous sociodemographic variables and substantially mediated by the health/functional limitation status measures.

Results indicate that service-connected disability status accounts for some variation in, and may have a cumulative effect on, the veteran mortality disadvantage. When possible, future research should account for service-connected disability status when studying veteran–nonveteran mortality differentials.