This quantitative research examined the variables affecting the Turkish listening skills of high school students living in Moldova-Gagauzia. It was evaluated effects of language exposure, linguistic distance, demographic variables (gender, settlement type, mother tongue, and Turkish language level) on participants' Turkish listening skills. 148 participants were students at Süleyman Demirel Moldo-Turkish High School in Kongaz. The data were collected using two instruments: The Turkish language exposure scale and listening tests. The scale was developed by Çobanoglu Aktan and Inan (2020) to determine Turkish language learners' out-of-school exposure to Turkish. Listening test was developed based on Yunus Emre Institute's "Yedi Iklim Türkçe Ögretim Seti". Afterward, the listening test results were analyzed to determine the variables that predicted listening performance. The research results indicated a low level of Turkish exposure among the participants. Females were exposed to Turkish more. There was no relationship between their listening test performance and exposure, mostly including Turkish visual and auditory media. In addition, in the Turkish listening exam, it was seen that the native speakers of Gagauz were more advantageous than Russian. The reason why the A1-2 level participants were more successful in the listening skill test might be the linguistic proximity and similarities of basic vocabulary between Turkish and Gagauz. It was concluded that linguistic distance was a stronger predictor than other variables.