Original version television represents not only a popular and accessible source of entertainment but also a valuable tool in foreign language learning and teaching. Building upon the potential of verbatim captions to facilitate speech segmentation and auditory word recognition, this study investigated pronunciation learning through exposure to video clips from TV series. The participants were three groups of Spanish/Catalan high school learners with an intermediate level of English (N = 53). Over 6 weeks, two groups watched videos either with textually-enhanced captions (n = 18) or unenhanced captions (n = 17) and engaged in pronunciation-focused audiovisual activities, targeting regular past <-ed> pronunciation. A control group (n = 18) of learners who were not exposed to the intervention materials also underwent pre- and post-testing. We used a read-aloud task and a narrative task based on picture prompts to assess past <-ed> pronunciation accuracy in controlled and spontaneous speech, respectively. In addition, we controlled for individual differences in English vocabulary size, listening comprehension skills and phonemic coding ability. The results revealed that past <-ed> pronunciation accuracy in controlled speech did not significantly improve at post-test. However, the group exposed to unenhanced captions exhibited significant pronunciation gains in the post-test narrative task, which were maintained at delayed post-test. The factors contributing to the successful implementation of the intervention and the potential challenges encountered by the enhanced group are discussed.