As Tarone (1998) stated, an understanding of interlanguage variation in relation to contextual changes has been a key issue in both second language acquisition (SLA) and language assessment (LA) research. Research on interlanguage variation has shown that systematic variation is often evidenced when different phonological and syntactic forms are examined across contexts. Such systematic variation has raised important questions about the distinction between competence and performance, and generalizability of results beyond the research elicitation tasks or test tasks in SLA and LA research. However, most previous studies that have examined this issue are based on cross-sectional data with a focus on between-group differences rather than within-individual differences across different contexts. Such limited data often make it difficult for researchers to understand individual developmental trends in interlanguage as well as to interpret context effects on the learner's spoken and written language data. Electronic (E)-portfolios may address this limitation by serving as a valuable research and assessment tool for collecting and storing an individual learner's language samples obtained across different tasks over time. The technology may also enhance the situational and interactional authenticity of tasks by including multi-media input and constructed response tasks. However, it is not clear how tasks in E-portfolios can be constructed to represent various linguistic and situational contexts, and how they could be systematically evaluated and scored. In that vein, this article addresses a number of limitations of existing E-portfolios as a research and assessment tool, and offers recommendations and suggestions for future research.