Abstract: How do temporary spells out-of-employment affect individuals’ labor trajectory? To answer this question, we leverage a 50% employer-employee matched dataset from Hungary for years 2009-2017, that also contains the entire history of individuals’ drug prescriptions and medical diagnosis. We use unexpected accidents (e.g., fracture of arm, or dislocations of joints) with arguably no permanent productivity impacts, to identify the effect of random assignment to temporary non-employment. Using both matching and instrumental variable identification strategies and controlling for a rich set of individual- and firm-level characteristics prior accidents, we find that, after non-employment spells following temporary accidents, individuals’ wages decrease substantially for up to two years. We show that most of the wage effect is due to reallocation of workers to lower quality employers (captured by firm size, ownership structure and AKM firm effect). We interpret our results through the lens of an augmented “job ladder” model, in which individuals receive alternative take-it-or-leave-it wage offers from firms and potentially suffer accidents which may push them into the state of non-employment. In such an environment, conditional on not suffering an accident, wages grow via job switching (both within and across employers). In a non-employment spell following an accident, the individual becomes less selective on which wage offer to accept relative to when she had a job. Thus, upon receiving a health shock and not having the option to return to their previous employer, the individual “falls from the job ladder”. However, even if the individual is able to return to the same exact job after recovery, her wage is smaller than in the counterfactual of no accident; the intuition is that she did not have the opportunity to receive alternative job offers that she would have received had she remained employed, and thus foregoes the opportunity to “climb the job ladder”. Both mechanisms alter the entire labor trajectory of individuals, with persistent wage and reallocation effects.