This study aims to follow up a recently performed community study of 624 grade 3 pupils who 2012 completed assessments of working memory, test anxiety, and self-regulation, and who then took the national test in mathematics. In the present study, results of national tests and grades in mathematics will be collected. The study will add four additional time points (when pupils are in grade 6, 7, 8 and 9, respectively) providing a longitudinal perspective of the relations between psychoeducational measures and mathematical achievement, in a Swedish school sample. It will therefore be possible to investigate the predictive power for mathematical achievement. The study signifies the need to identify risk factors for school failures, and ultimately develop instrument that can be prognostic of school attainments in general, and mathematics in particular.