We analyze the effectiveness of the early childhood programme (ECP) in the Netherlands. The programme is designed for 2.5-4 year olds from disadvantaged backgrounds. 37 municipalities received an additional subsidy to expand ECP programmes, which allows us to analyze the effects of the programme within a difference-in-difference-in-differences framework. Most children first enroll in primary schools at age 4 in the Netherlands, but pupils begin to learn reading and mathematics in grade 3 at age 6. We use grade repetition constructed from school registry data from 2008 to 2015 in the first two grades as an indicator of school readiness. Our results show significantly lower grade repetition rates for targeted boys who are in regions that receive the subsidy. Grade repetition drops by 0.8-1.8 percentage points from a mean of 10.5% for the disadvantaged group targeted by the programme. (author abstract)