Heritage language learners are individuals who enroll in classes to (re)learn their family language in an instructed setting. Their family or heritage language is a minority language in the society at large. As such, they are (multi-)bilingual speakers of the heritage language(s) along with the majority language(s) of the society. The collaborative efforts from scholars and practitioners from different disciplines (e.g., sociolinguistics, applied linguistics) have been essential to better understand and serve the needs of these learners. This has led to the field of heritage language education. Within this broad field, however, I will focus on research that has investigated the effects of instruction on heritage language learners’ language development and performance. The first part of the talk will consist in broadly framing the key issues at stake with regard to heritage language learning outcomes, which includes the results of a recent exploratory meta-analysis. During the second part of the talk, I will share quantitative and qualitative data from three empirical studies under a task-based language learning framework that shed light into how heritage language learners respond to instruction. I will conclude the talk with a few take-away messages based on task-based research with heritage language learners that have implications for optimizing pedagogical decisions.