This paper draws on data from TIMSS 2019 to investigate an approximate relation between Danish, Swedish, and English teachers’ in-service training in computational thinking (CT) and their students TIMSS scores as compared to students whose teachers had only received training in digital technology (DT) or problem solving (PS). To that end, we developed a proxy for CT combining PS and DT, based on recent definitions of CT in mathematics education. The results indicate that, in England, in-service CT training is associated with systematically better student TIMSS scores as compared to students whose teachers received only DT or PS training, but the best scores were achieved by teachers who did not participate in any of the three forms of training. In Denmark, CT training was found to have a negative effect on TIMMS scores while in Sweden, the effect of CT training was difficult to assess.