This study analyzes the effect of all-day (AD) primary school programs on maternal labor supply. To account for AD school selectivity and selection into AD primary school programs I estimate bivariate probit models. To identify these models I exploit variation in the allocation of investments to AD primary schools across time and counties. This variation results from the public investment program "Future Education and Care" (IZBB) which was introduced by the German federal government in 2003. My results indicate for mothers with primary school-aged children in Germany (excluding Bavaria) a significantly positive effect of AD primary school programs on labor supply at the extensive margin. On average, mothers who make use of AD primary school programs are 26 ppts more likely to be employed than mothers who do not make use of these programs. This large effect is robust to alternative specifications and estimation methods and mainly concentrated in states with AD primary school student shares of up to 20%. On the contrary, there is no evidence for an impact of these programs on maternal labor supply at the intensive margin (full-time vs. part-time).