BackgroundIn the past years, cumulative evidence has convincingly demonstrated that the work environment is a critical determinant of workers' mental health. Nevertheless, much less attention has been dedicated towards understanding the pathways through which other pivotal life environments might also concomitantly intervene, along with the work environment, to bring about mental health outcomes in the workforce. The aim of this study consisted in conducting a systematic review examining the relative contribution of non-work determinants to the prediction of workers' mental health in order to bridge that gap in knowledge.MethodsWe searched electronic databases and bibliographies up to 2008 for observational longitudinal studies jointly investigating work and non-work determinants of workers' mental health. A narrative synthesis (MOOSE) was performed to synthesize data and provide an assessment of study conceptual and methodological quality.ResultsThirteen studies were selected for evaluation. Seven of these were of relatively high methodological quality. Assessment of study conceptual quality yielded modest analytical breadth and depth in the ways studies conceptualized the non-work domain as defined by family, network and community/society-level indicators. We found evidence of moderate strength supporting a causal association between social support from the networks and workers' mental health, but insufficient evidence of specific indicator involvement for other analytical levels considered (i.e., family, community/society).ConclusionsLargely underinvestigated, non-work determinants are important to the prediction of workers' mental health. More longitudinal studies concomitantly investigating work and non-work determinants of workers' mental health are warranted to better inform healthy workplace research, intervention, and policy.