In France, childless adults younger than 25 face hard-to-meet eligibility conditions to enroll in the minimum income program. The restrictive requirements generate a “jump” in the number of recipients at ages around 25. We use a Regression Discontinuity (RD) design to assess the impact of the French minimum income program (RSA) on users of accommodation and meal distribution services. We find that the RSA benefit reduces the homelessness rate by 20% among young adults aged 22 to 27. This result is driven by new RSA recipients who have started paying partial rent to third parties, and the probability of becoming a regular tenant increases after age 26. We simulate the effects of lowering the program’s minimum age eligibility on the probability of being homeless. Our findings suggest that in programs directed at homeless individuals, around 60% of expenditures are offset by savings in social assistance costs to the homeless.