Although the effectiveness of cadaver surgical training has been clarified, the decline in training effects over time has become a problem. This study examined whether repeated participation in cadaver-based educational seminar for trauma surgery (C-BEST) could suppress the decline in training effects. Basic and advanced C-BEST have pelvic package (PP) and fasciotomy of the lower extremity (FLE) as common training skills. For participants of these skills twice each, we examined the changes in a 10-point self-assessment of confidence levels (SACL) at six time points: (1) before the seminar of basic C-BEST, (2) immediately after basic C-BEST, (3) half a year after basic C-BEST, (4) before advanced C-BEST, (5) immediately after advanced C-BEST, and (6) half a year after advanced C-BEST. Data were collected from 28 basic C-BESTs and 5 advanced C-BESTs conducted from January 2013 to January 2020. Statistical analysis was performed by comparing SACL results from seminar evaluations at the six points, with significance at P < 0.05. A total of 60 participants were enrolled (postgraduate year, 16.5 ± 5.7). The interval between basic and advanced C-BEST was 27.1 ± 6.9 months. In PP, the SACL did not decrease at all six points. In FLE, SACL did not decrease at all six points, had a greater increase before versus immediately after advanced C-BEST, and did not decrease thereafter (P < 0.05). After participants retook the seminar, FLE-like procedures, which are unfamiliar to nonorthopedic surgeons, had increased and maintained self-evaluation values, whereas PP-like procedures, which are familiar to abdominal surgeons, had maintained high self-evaluation values. Therefore, repeated seminar participation could maintain the effects of cadaver training.