Over the past few decades, the field of second language acquisition (SLA) has seen a remarkable increase of interest in the study of instructed second language acquisition (ISLA), which "investigates second language (L2) learning or acquisition that occurs as a result of teaching" (Loewen, 2014, p. 2). Research insights gained from this subfield are particularly pertinent to adult L2 learners, who, due to biological and cognitive constraints, have difficulty acquiring a target language (TL) solely based on naturalistic input (e.g., Han, 2004; Long, 1990). The ISLA literature shows that there is an array of pedagogical options that can be used to facilitate adult L2 learning, ranging from implicit (e.g., input enhancement, recasts, etc.) to explicit (e.g., consciousness-raising, metalinguistic rule explanation, etc.) techniques. Furthermore, the effectiveness of an instructional treatment seems to depend largely on the nature (i.e., complexity) of the L2 feature (e.g., Ellis, 2002; Spada & Tomita, 2010). However, extant empirical studies have yielded rather mixed findings on the issue regarding which type of L2 feature (i.e., complex or simple) benefits more from which type of instruction (i.e., implicit or explicit), rendering it difficult to provide straightforward guidance to L2 classroom teachers. There are several reasons for the disparities in research findings, such as differences in study designs, settings, learner characteristics, etc., but above anything else, the inconsistent findings can primarily be attributed to the varying conceptualizations of "complexity." With an aim to enlighten future research in this line of inquiry, the present discussion emphasizes the need for a more integral definition of complexity. First, some traditional definitions of the concept are briefly reviewed. Next, a more recent, acquisitional perspective (Han & Lew, 2012) is introduced, and finally, a few key aspects of "acquisitional complexity" are discussed, which offer critical insights for future empirical studies, particularly related to the internal validity of research designs.