In this paper we discuss the rationale for tracing non-original sample members (Non-OSMs) in household panel studies, and in particular in SOEP, and the implications for weighting. We present results on the incidence, survival rates, and thus the relevance of Non-OSMs in the SOEP.