In the paper we define, elaborate and illustrate what language documentation is, firstly by focusing on the global context where the large number of vanishing languages warrants such an approach in applied linguistics, and then by narrowing our scope to the local context of the Republic of Serbia, which itself has a certain number of endangered and vulnerable languages and language varieties. Indicating the discrepancy between official records found in world catalogues and databases (UNESCO’s Atlas of the World’s Languages in Danger, Ethnologue, and The Catalogue of Endangered Languages (ELCat)) and the situation in the field as described by researchers familiar with the local context, we investigate the experiences of a team of researchers currently working on documenting Aromanian, Banat Bulgarian, Vojvodina Rusyn (Ruthenian), Judezmo (Ladino), Romani (Gurbet and Arli varieties), Megleno-Romanian and Romanian (Bayash and Vlach varieties) as endangered and vulnerable languages and language varieties in Serbia. Using interviews as a method for collecting data, conveying the researchers’ experiences in the field and their ways of dealing with various issues that come up in the process of language documentation, we have singled out two broad fields of topics. The first one tackles methodological challenges encountered prior and during field work, whereas the second focuses on ethical issues that necessarily must be resolved due to the participation of human subjects. In terms of the first group of topics, we have focused on research design (considerations involved in the project preparation), sampling (usually relying on researchers’ previous experiences more than on official records), personal involvement (often inevitable, but researchers do present strategies that help them maintain objectivity), and establishing first contact and recruiting informants (which also relies on previous work in the community). As for the second group of topics, that concerns ethical issue, the researchers have discussed the ways of getting informants’ consent for research and how they explain the importance of language documentation for that particular community and its members. Additionally, it addresses the issue of compensation that sometimes occurs, where informants expect either material or other kind of reward for helping out the research team. The paper also discusses how researchers deal with various sensitive topics that may range from personal issues to political opinions, and finally, the ethical responsibility that the researchers have towards the community they are studying in terms of giving something back and providing a sustainable relationship towards the language in question. We conclude by positioning the process of language documentation in the wider field of social studies, indicating its importance not only for sociolinguistics but also for the sociology of language, sociology of ethnicity and other related disciplines.