Recent network research indicates that native youth prefer to befriend immigrants with stronger rather than weaker host country identification. Surprisingly, however, no respective preference of high-identifying immigrants for native friends has been found, and there is little evidence that friends influence immigrants' identification. Seeking to make sense of these unexpected findings, my aims are twofold: First, I reproduce an earlier study using three waves of newly collected network panel data. Second, going beyond a robustness test with better data, I suggest that relative group size within school accounts for earlier findings. I hypothesize that immigrants' host country identification only affects their own friendship choices in schoo...