Understanding abstract concepts and ideas in mathematics if instruction takes place in the first language of the student is challenging as it is. Yet worldwide students often have to master mathematics via a second or even a third language and this is a recognised problem. The majority of students in South Africa – a country with eleven official languages – have to face this difficulty. What is the extent of the linguistic disadvantage (if any) of South African second language students? With the language challenges that South Africa faces today with respect to education as a backdrop, I report on a quantitative investigation into this problem specifically for the case of tertiary mathematics students, focussing on Afrikaans first language learners. The performance of Afrikaans first language students who attend Afrikaans lectures is compared to that of the Afrikaans first language students who attend English lectures. In a further investigation I explore the influence that second language instruction has on students from African cultures. All the students study at the University of Pretoria. The study shows that in the comparison of the average performances of the two Afrikaans first language groups, there is no significant difference in the first year university calculus performances, but at secondary school level the Afrikaans students who attend English lectures at university level achieve better results. When the means are adjusted by removing the influence of school mathematics achievement, the adjusted average performance of the Afrikaans group that attend English lectures differ significantly from that of the Afrikaans group that attend Afrikaans lectures. As a result both of the analyses suggest that, based on mathematics achievement, Afrikaans students who attend English lectures may be at a disadvantage to Afrikaans students attending English lectures do. The study also indicates that the African students' performances do not differ significantly from that of the Afrikaans students who attend English lectures (both of these groups attend second language lectures). In the comparison of the pooled groups of first language learners and second language learners, there does not seem to be any significant difference between the adjusted mean performances of these groups.