This study explores whether explicit knowledge and production of present continuous are equally affected by a task-based peer-interaction intervention with and without pre-task grammar and interactional instruction. Four groups of 4th-grade Catalan-Spanish bilingual EFL learners, aged 9–10, participated in an 8-week pedagogical intervention (50 min per week), while a fifth control group only participated in the testing sessions. Groups were divided according to the type of pre-task instruction received on a weekly basis: Grammar + Interaction Strategy (n = 22), Grammar (n = 23), Interaction Strategy (n = 22), and Task-Only (n 23). The four groups also participated in task-based peer interaction every week (both collaborative and information-gap tasks). All groups took part in a bi-modal, untimed Grammaticality Judgement Task (GJT) and carried out a collaborative picture narrative task before and after the intervention. In the GJT, the analysis revealed that only those groups who received either mixed grammar-interaction or grammar-only instruction showed significant improvement in accuracy, modulated by condition. In the production task, the analysis showed that the four treatment groups exhibited an increase in both production attempts and accuracy and differed from the control group after the intervention, though only the groups that received pre-task grammar instruction improved significantly. Results will be discussed in relation to the pedagogical intervention and the development of explicit knowledge and production of present continuous.