In Norway, immigrants receive higher levels of social assistance than natives. How can we explain this difference? After controlling for differences in take-up rates through a two-step Heckman procedure, we attempt to answer this question by exploiting rich data from administrative registers. We operationalise social assistance in the Norwegian context by employing a composite variable that includes: (a) financial assistance, (b) housing allowance and (c) qualification benefit. We quantitatively analyze the difference in benefit levels of social assistance between the first and second generations of immigrants and the benchmark levels of the non-immigrant population through a Kitagawa-Oaxaca-Blinder (KOB) decomposition exercise. The results of the analysis indicate that a significant portion of the gap in benefit reception between immigrant groups and natives is due to observable characteristics (42% for immigrants and 69% for their descendants), with unobservable cultural and behavioural factors explaining the remaining portion of the gap.