Preparing students for the 21st century means teaching them higher-order thinking (HOT). This requires a thorough theoretical and practical understanding of how HOT develops and can be fostered. This Interlinked Research Project (IRP) aims to contribute to this understanding by taking a domain-specific conceptualization of HOT and focusing on primary mathematics education. While internationally, awareness is increasing that the foundation of HOT in mathematics has to be laid at young age, in Dutch primary schools it is almost absent. To provide evidence for enriching the current mathematics curriculum the IRP investigates in three part-projects what opportunities-to-learn HOT can be offered by dynamic data modeling, probability, and early algebra. The IRP is grounded in embodied cognition theory, representational re-description theory, and variation theory and uses interventions with ICT. The design combines macro-genetic longitudinal studies (development over one school year) and micro-genetic studies nested within the macrogenetic studies (development over lesson series).