Marginal employment (ME) is one of the largest forms of atypical employment in Germany. We analyse whether ME has a ‘stepping stone’ function for unemployed individuals, i.e., whether ME increases the subsequent probability of regular employment. We find differing treatment effects by unemployment duration. According to our results, ME increases the likelihood of regular employment within a 3-year observation period only for those who take up ME several months after beginning to receive benefits. In contrast, for those starting ME within the first months of receiving benefits, there is no effect on the probability of regular employment. Although we took several measures to minimize the impact of unobserved heterogeneity, our results can only be interpreted as causal if the conditional independence assumption holds.