To address the challenges of limited language proficiency and provide necessary feedback in the implementation of task-based language teaching (TBLT), a mobile-supported TBLT application was developed to provide linguistic and task scaffolding. Sixty-six English as a foreign language (EFL) university learners participated in a three-week experiment as part of a general English course. They were assigned to either an experimental group (mobile-supported TBLT), which received TBLT with scaffolds built into the application, or a control group (traditional TBLT), which received traditional paper-based TBLT without the scaffolds. At the end of the experiment, an English achievement test of vocabulary, grammar, and conversation comprehension was administered to determine if the technological scaffolds enhanced the learning outcomes for the course. Students’ self-perceived use of oral communication strategies was also measured to explore how these scaffolds affected the conversational interaction essential for task performance. Results showed that the mobile-supported TBLT group outperformed the traditional TBLT group on the vocabulary and conversation comprehension tests but not so much on the grammar test. Also, the mobile-supported TBLT group reported greater awareness of fluency- and accuracy-oriented strategies for speaking than the traditional TBLT group. Implications for designing mobile learning to enhance TBLT in an EFL setting are drawn.