This paper presents a scientometric analysis of task-based teaching and learning using CiteSpace based on the research literature retrieved from the Web of Science core collection between 2013 and 2022. CiteSpace is employed to analyze the data, as it minimizes subjectiveness and increases credibility. A total of 1799 bibliographic records related to task-based teaching and learning were systematically visualized and examined to identify the themes, knowledge evolution, and emerging trends in this field. The results show that five major themes can be identified, including task-based language teaching, instruction, computer-mediated communication, design, and working memory and self-efficacy. Knowledge evolution of task-based teaching and learning experienced three phases of development. Through the burst detection of keywords and cited references, and examining the major citing references of the detected keywords and references, it is found that the emerging trends focus mainly on technology-assisted learning, task design, task evaluation, etc. The findings provide insights for future researchers and educators.