The importance of investments in early childhood education (ECE) has been widely documented in the literature. Among the benefits, particularly for children from disadvantaged backgrounds, is its potential to mitigate educational inequality. However, some evidence also suggests that the positive effects of ECE on later outcomes tend to dissipate over time, leaving children who attended such programmes no better off academically than those who did not. This paper studies the relationship between students’ years spent in ECE, from 0 to before starting primary school, and the results of their educational assessment outcomes at age 15. Using PISA survey data for 14 European countries from 2015 to 2018, we conduct a cross-country comparison of student performance in reading, mathematics, and science, correlating the results to the duration of ECE attendance. Our findings show that duration in ECE is associated with better assessments at age 15, but that the benefit is nonlinear and peaks at 3–4 years of attendance. Gender and migration background are associated with student performance on the assessments; but we don’t find evidence of heterogeneity in the relationship between ECE duration and test outcomes based on gender and migration background. Instead, we document differential effects of ECE duration according to age of entry to ECE, mother’s education, and the type of educational system attended.