Task-based language learning and teaching research from both psycholinguistic and pedagogical perspectives shares a common theoretical background of learners’ attention, awareness, and perception (Levelt, 1989). The former has focused on learners’ prioritized attention to language aspects (e.g., fluency) in their oral performance. Furthermore, researchers have explored learners’ attention during strategic planning through emergent categorization, from retrospective data (e.g., Ortega, 2005). The latter has focused on learners’ uptake, based on incorporation from teachers’ corrective feedback (e.g., Mackey & Philp, 1998). The underpinning concept of incorporation via noticing a gap in Schmidt (1990) displays learners’ awareness of linguistic factors. The present study attempts to fill a gap in previous research by employing incorporation as a more reliable measure, of learners’ attention to linguistic factors, than retrospective data. Allocation of attention in four learners’ oral performance is qualitatively explored over five task repetitions by employing emergent categories of linguistic incorporation. This reveals what learners do during planning in their oral performance and how allocation of their attention changes across five task repetitions. This has long been a puzzle in quantitative analysis of such data. The students’ linguistic incorporation demonstrates their attention to different linguistic factors (e.g., semantic, syntactic), which are linked to a priori categories of fluency and complexity in their oral performance. This allocation of attention eventually changes over task iterations. The trajectories of fluency and complexity are also confirmed by supplemental examinations of data from 15 students. This suggests that individual learners prioritize their attention to a particular area (Foster & Skehan, 2013), and then broaden attention to other areas as more space becomes available for processing through repeated use (Bygate & Samuda, 2005). Besides this cognitive demand, the present study also reveals that learners’ attention may be affected by interlocutor familiarity, social, and cultural factors.