Crossover effects of critical life events within families have received growing attention in life-courseresearch. A parent losing a job is among the most distressing events that can befall a family, butexisting research has reached discrepant conclusions concerning if, and if so how, this affects childmental health. Drawing on insights from models of intra-family influence and life courseepidemiological models, we ask if parental job loss have latent or long-term effects on child mentalhealth, if the effects are conditional on the timing of the job loss, and if repeated job losses havecumulative effects.We use intergenerationally linked Swedish register data combined with entropy balance andstructural nested mean models for the analyses. The data allow us to track 400,000 children over 14years and thereby test different life-course models of crossover effects. We identify involuntary joblosses using information on workplace closures, thus reducing the risk of confounding.Results show that paternal but not maternal job loss significantly increases the risk of mental healthproblems among children, that the average effects are modest in size (less than 4% in relativeterms), that they materialize only after some years, and that they are driven by children aged 6-10years. Moreover, we find evidence of cumulative effects, but also of declining marginal harm ofadditional job losses over the life course.