Theory and evidence point towards particularly positive effects of high-quality child care for disadvantaged children. At the same time, disadvantaged families often sort out of existing programs. To counter differences in learning outcomes between children from different socioeconomic backgrounds, European governments are pushing for universal child care. However, evidence on the effects of universal programs is scarce. We provide evidence on the long-run effect on schooling of mandating kindergarten at age 5–6. Our identifying variation comes from a reform that lowered school starting-age from 7 to 6 in Norway in 1997. Our precise DD estimates reveal hardly any effect, both overall, across subsamples, and over the grading distribution. A battery of specification checks support our empirical strategy.