The results of studies targeting cognitive and academic advantages in children frequenting early bilingual immersion school programs (CLIL) have been contradictory. While the impact of the amount of CLIL experience has already been studied, the role of the second language learned has been little studied to account for differences among study findings. The link between executive skills (EF) and scholar abilities (e.g., mathematics) in the CLIL context has also been little investigated. The purpose of the present study was to determine if the impact of CLIL on EF and academic performances varies depending on the immersion language and the duration of CLIL experience. The sample included a total of 230 French-speaking children attending second (141) and fifth (89) grade classes. Within each grade, there were three matched language groups composed of children respectively immersed in English, immersed in Dutch, and non-immersed controls. The children were administered tasks assessing executive functions [alerting, cognitive flexibility, and working memory], as well as arithmetic abilities. In second grade, we detected no difference in EF between the language groups. On the other hand, in fifth grade, the two immersed groups outperformed the non-immersed group on the cognitive flexibility task but did not differ between them. Moreover, only the Dutch immersed group outperformed the control group on the working memory task. Arithmetic performances also differed depending on the language learned; in second grade, Dutch learners performed better than the monolingual group. In fifth grade, Dutch learners outperformed the two other groups. These results suggest that the impact of CLIL on executive skills and arithmetic performances might be modulated by the amount of CLIL experience and the second language learned in immersion.