Previous literature highlights the potential of ICT use to enhance mathematical learning. There are also several theoretical arguments supporting that gifted education benefits from ICT use. However, empirical studies have paid little attention to the relationship between ICT use and gifted students' mathematics performance. It is also unclear whether and why this relationship differs between gifted students and their peers. For the first time, we tested this relationship by using a large-scale multinational sample of 236,938 adolescents attending 10,213 schools in 44 countries in several contexts from the Programme for International Student Assessment Questionnaire (PISA) 2018. We estimated a hierarchical linear model (HLM) and found that only gifted students benefit from ICT use in mathematics learning. The higher their level of performance, the more beneficial ICT use is for gifted students. This relation is negative in the case of the rest of the students. Based on theoretical arguments, we also explain the likely reasons that lay behind this different relationship between gifted students and their peers. The findings illustrate that policymakers should consider a differentiated approach to ICT use at school depending on the students' level of performance. Gifted students could benefit more from ICT use in learning and the rest of the students from teaching with more human interaction.