The aim of this study was to investigate the effects of the dialectical constructivist instruction and portfolio-based assessment on 219 primary-school pupils' reading comprehension achievement. Pupils were assigned to the intervention condition (experimental classes) or to a traditional instruction condition (control classes). The pupils of the experimental classes were taught with the dialectical constructivist instruction characterised by four reading strategies (summarising questioning, clarifying, predicting) and were assessed according to a portfolio-based approach during the experimental period. Compared to conventional reading lessons (the control class without treatment), the intervention of the dialectical constructivist instruction enhanced pupils' mastery of reading comprehension competencies. Furthermore, the portfolio-based assessment employed to generate pupils' authentic performance were outperformed by the experimental group at the post-test.