Using data from the mid-1990s from the Luxembourg Income Study (LIS), we analyze for several household types the effect of mothers' work participation on families' relative income position and poverty risk. Results are compared across seven European countries with contrasting family policies: the UK, Germany, France, the Netherlands, Italy, Sweden, and Finland. Findings based on traditional research designs first indicate a strong positive effect of mothers' labor market participation on families' income situation in almost all countries and family types. By applying selectivity models, however, we show that a substantial part of this positive effect is caused by the fact that mothers in employment are a positively selected group. This implies that continuing policy measures to expand mothers' labor market participation - which are, nevertheless, strongly recommended - are likely to become less efficient as the work participation of mothers increases.