This paper analyses the effects of the size of school districts on educational attainment of students after compulsory school in Denmark. Using administrative microdata for individual students and their parents, logit models for the probability of obtaining different levels of education are estimated where the explanatory variables include district size, per student expenditure, and a wide range of controls for family background and district socioeconomic characteristics. It is found that attending schools in districts with a population of less than 15,000 (corresponding to less than 1800 students in public primary and lower secondary schools) has negative effects on educational attainment later in life: The probability of completing a secondary education (upper secondary school or vocational education) or a further or higher education is reduced by approximately 3 percentage points compared to having attended primary and lower secondary school in a district with more than 15,000 inhabitants.