Learning styles (LS) have dominated educational practice since their popularization in the 1970s. Studies have shown that they are accepted by more than 90% of teachers worldwide. However, LS have also received extensive criticism from researchers and academics, due to the poor theoretical justification of the theory, their problematic measurement, and the lack of systematic studies supporting them. The present study tested the hypothesis that teachers' and students' assessment of preferred LS should correspond. Moreover, it tested whether teachers' judgment of LS is driven by the students' IQ. Both questions were studied for the first time in a systematic fashion within LS research in primary school pupils. Fifth and sixth grade pupils (n = 199) were asked to self-assess their preferred LS, while their teachers were asked to provide their own assessment on individual pupils' LS. No relationship was found between pupils' self-assessment and teachers' assessment, suggesting that teachers cannot assess the LS of their students accurately. Moreover, students' intelligence was not found to drive teachers' assessment of their LS. This study adds to the body of evidence that is skeptical of the adoption of LS in mainstream education.