Purpose: To explore the relationship between problem-based learning (PBL) evaluations and clinical performance. Methods: The study included 117 third-year medical students at Chonnam National University Medical School. The students' first-, second-, and third-year PBL evaluations were compared with their clinical performance examination (CPX) scores in third year. The PBL evaluations were composed of three subscales: tutors' evaluation (students' professional behavior, contribution to group process, and contribution to group content), a report, and a written examination. The CPX assessed four performance categories: history taking, physical examination, information sharing, and patient-physician interaction. Results: No significant correlation was found between the first-year PBL evaluations and CPX scores; however, the second-year PBL evaluations were significantly correlated with history taking (r=0.186, p=0.044) and patient-physician interaction (r=0.213, p=0.021) of CPX. The third-year PBL evaluations were significantly correlated with physical examination (r=0.248, p=0.007), and patient-physician interaction (r=0.283, p=0.002) of CPX. Several significant correlations between the PBL evaluations subscales and CPX scores were revealed. The PBL tutors' evaluation (r=0.343, p=0.000) and report scores (r=0.210, p=0.023) were significantly correlated with patient-physician interaction of CPX. The contribution to group process of tutors' evaluation was significantly correlated with patient-physician interaction (r=0.186, p=0.045), and the contribution to group content of tutors' evaluation was significantly correlated with physical examination (r=0.187, p=0.044). Conclusion: We found a significant association between PBL evaluations and CPX scores.