The testing effect, defined as the positive effect of "retrieval practice" (i.e., self-testing) on long-term memory retention relative to other ways to support learning, is a robust empirical phenomenon. Despite substantial scientific evidence for the testing effect, less is known about its effectiveness in relation to individual differences in cognitive ability. In the present study, we examine whether the effect of retrieval practice is beneficial independent of cognitive ability using behavioral and brain imaging data. In a within-subject design, upper-secondary students learned Swahili-Swedish word pairs through retrieval practice and study. The testing effects were assessed at a direct test and for a subsample after 1- and 4-weeks retention intervals, respectively. Another subsample performed the 1-week retention test during functional MRI (fMRI). Memory retention was analyzed in relation to an educationally relevant composite score dividing participants into low, intermediate, and high cognitive-ability groups. We provide behavioral evidence that the testing effect is independent of cognitive ability. The fMRI findings confirmed a general effectiveness of retrieval practice by showing that brain regions associated with successful retrieval of conceptual representations and semantic processing were more strongly engaged after retrieval practice in all cognitive-ability groups. It is argued that the advantages of retrieval practice should be conveyed to all teachers and students.