In Content and Language Integrated Learning (CLIL), the design of a language policy at school level is not only bound up with the number of languages used for learning and the number of hours devoted to teaching those languages, but also with the fact that language becomes specialised in relation to the subject, which impacts on the methodology used. These are the reasons for both language teachers and subject teachers to work together in design and implementation; and for the teachers' use of a translanguaging-based approach to language learning (San Isidro, 2018). Previous research has dealt with teachers' opinions (Calvo & San Isidro, 2012; Coonan, 2007; Infante et al., 2009; Pladevall-Ballester, 2015) on the difficulties of curriculum integration and its effects on both the different languages of instruction and the learning of content; or on the difficulties of language and content integration. However, methodology-oriented research on teachers' views and work in specific contexts is direly needed so as to gain a deep insight into the methodological commonalities that make CLIL what it is. Our qualitative study is focused on a two-year monitoring of teachers' (N=6) views on CLIL implementation in a rural multilingual setting in Galicia. The teachers were monitored by means of interviews held between 2012 and 2014. After being trained, they took part in a CLIL project based on curriculum integration with two different groups of students. The findings reported showed that 1) teachers' initial views on CLIL implementation turned more positive over the two years; 2) teachers believed that CLIL provides a very good framework for the development of pluriliteracies; 3) their views regarding content learning in CLIL turned more neutral in the course of the two years; and 4) teachers stressed the need for methodology-oriented training.