This paper points to the theoretical and methodological issues involved in the subproject on Portuguese Language Atlas in Indigenous Areas (ALiPAI) part of the project GeoLinTerm coordinated by Abdelhak Razky (UFPA/UnB). The first results of ALiPAI are part of the doctoral thesis of Guedes (2017), which mapped the geosociolinguistic profile of Portuguese in contact with Tupí-Guarani languages in indigenous areas of Pará and Maranhão. The first experiences in the geolinguistic field research in Brazilian indigenous areas provided a confluence of research methodologies, especially Geosociolinguistics (RAZKY, 1998), Pluridimensional and Relational Dialectology (THUN, 1998) and Anthropological Linguistics (RODRIGUES, CABRAL, 2012). Ten informants were selected from each of the five network of points investigated (Suruí Aikewára, Asuriní do Tocantins, Tembé, Guajajára and Guaraní Mbyá). The questionnaires applied in these areas were developed by the National committee of the Linguistic Atlas of Brazil - ALiB: PhoneticPhonological questionnaire (QFF) and Semantic Lexical Questionnaire (QSL). These questionnaires were adapted to include the correspondence request in the indigenous language for each one of the answers obtained in Portuguese. In addition, a complementary QFF and a Sociolinguistic Questionnaire were used. The paper also reflects on the necessary adaptations made in the methodology of geolinguistic research, to account for the geosociolinguistic characteristics of the ALiPAI target communities.