The aim of this paper is to look into repairs in simultaneous interpreting in relation to directionality. The authors tried to test some hypotheses in the light of the literature available and the findings of an experiment designed specifically for this purpose and carried out in Vilnius, Lithuania, and Istanbul, Turkey, with a total of eight students at two interpreting schools in the language pairs Lithuanian-English and Turkish-English. The authors have tried to find out whether directionality affects the types and frequency of repairs made by the interpreter, and whether the impact of directionality regarding repairs varies according to the language pair. Although it is difficult to extrapolate the results to the larger interpreting population, the authors have observed different repair patterns by direction and language. The most obvious results were that error repairs outnumbered by far any other type of repair in any direction; the number of stylistic repairs was significantly higher when working into the mother tongue; repairs were always more frequent when working into the A language; and the language pair influenced the type of repairs made. Some pedagogical implications have been drawn based on these findings. Further progress in this research field could be made by including speed as a variable, and comparing interpreting students with professional interpreters to see whether the more experienced use of strategies affects repairs.