This paper discusses and empirically tests the relations between language proficiency and national identification with Germany among first generation immigrants in Germany. It presents three theoretical arguments: (1) language proficiency positively affects national identification; and contrastingly (2) national identification positively influences language proficiency; and (3) there is a reciprocal relationship between both constructs. To test these potentially contradictory claims empirically, we utilize data on first generation immigrants in Germany measured in four waves (1997, 1999, 2001 and 2003) from the German Socio-Economic Panel (SOEP) study. Language proficiency is operationalized with the variable proficiency in German language. Hypotheses are tested using autoregressive cross-lagged structural equation models. Findings demonstrate an effect of language proficiency on national identification among immigrants in Germany. However, data provide no support for an effect in the other direction.