ABSTRACT This article investigates the extent to which policy variation matters for refugee integration, focusing on two parallel local integration programs varying in intensity and comprehensiveness and their effects on refugees’ economic and socio-cultural integration. We employ 3-wave panel data including detailed information on Syrian refugees from the moment they were granted asylum status in the Netherlands up to four years later. After accounting for treatment assignment bias using coarsened exact matching, random effects analyses showed a significant increase over time in refugees’ Dutch language proficiency, identification with the host community, self-reliance, participation in volunteer or paid work, and welfare independence. However, we did not find convincing evidence to suggest that refugees’ integration process developed differently depending on the integration program. We discuss possible explanations for our findings and advance policy interventions that aim to foster refugee integration.