This study aims to redefine the current TOPIK speaking evaluation criteria so that they can assess not only linguistic accuracy but also overall communicative competence, including pragmatic competence. To this end, speaking performance data were collected from six Korean language learners based on Tasks 2 and 4 of the TOPIK Speaking Test. A group of speaking assessment specialists was organized to conduct two rounds of rating. Through this process, the evaluators exchanged opinions on the existing criteria and developed final evaluation standards and directions for revision. First, it was concluded that the “content and task performance”category should include an assessment of whether the response content and expressions are appropriate to the given situation and context, thereby reflecting the learner’s pragmatic competence. Second, in evaluating the appropriateness of discourse situations, the criteria should be further specified by dividing them into two elements: content and task performance and linguistic resources. Third, in the delivery category, the evaluation standards for pronunciation, intonation, and speech rate should be made more detailed and explicit. As practical guidelines for applying the proposed criteria to actual assessments, the study proposes determining the weighting of each scoring category according to its relative importance and using a detailed evaluation rubric to enhance scoring feasibility and reliability. It is expected that this study will provide practical insights for instructors teaching TOPIK speaking and serve as foundational data for future research on the assessment of pragmatic competence in TOPIK speaking.