SUMMARY
This study explores whether semantically elaborated keywords facilitates learning of Latin roots. The freshmen of a medical school were the treatment group provided with the translation equivalents of the target roots supplemented with keywords. The control group (second year students of the same medical school) received only Turkish translations; the anatomy courses in their first year had also familiarized them to the roots incidentally. After a pretest and the subsequent presentation of the self-study materials, they were informed of a similar test they will take three weeks later. The students assumed their testing is over; however, fifteen weeks later a similar test was administered to assess in which group the decline might be lower. With t-tests and ANOVA, all the pairwise comparisons (pretest – first posttest, pretest – second posttest and first posttest – second posttest) indicated statistically significant differences acquired by the treatment group with implications for English for academic purposes.