PURPOSE: Endoscopic spine surgery has been widely performed to treat degenerative spinal diseases to mitigate the risks and complications associated with traditional open surgery. However, endoscopic procedures pose challenges, including a limited field of view and a restricted operating space, which can affect the surgeon's learning curve. This study aimed to evaluate the learning curves for unilateral biportal endoscopic spinal surgery (UBESS) and percutaneous transforaminal endoscopic surgery (PTES) by performing cumulative sum (CUSUM) analysis and to assess their efficacies in managing degenerative spinal diseases. PATIENTS AND METHODS: This retrospective cohort study included 100 consecutive patients who underwent PTES and 100 consecutive patients who received UBESS. CUSUM analysis was conducted to assess the learning curve, with cutoff points used to categorize the early and late phases. These two phases were analyzed in terms of differences in operative time, hospital stay, complications, and patient-reported outcome measures (PROMs). Additionally, PROMs between the PTES and UBESS groups, which were performed by the same surgeon, were compared. RESULTS: CUSUM analysis revealed that the operative time for PTES and UBESS decreased after 35 and 28 cases, respectively. Both early- and late-phase cases exhibited significant improvement in all PROMs postoperatively. Furthermore, PROMs did not differ between patients who underwent PTES and those who underwent UBESS. CONCLUSION: Both procedures achieved comparable clinical outcomes with low complication incidences. However, achieving proficiency in PTES required a learning curve of at least 35 cases, whereas that in UBESS required a minimum of 28 cases.