Objective: To evaluate the repeatability and reproducibility of central axial length and peripheral axial length measurements in children using the Swan 660 M. Methods: This cross-sectional study was conducted from August to September 2024. A total of 60 children (6-15 years; no significant ocular diseases) were recruited from Zhoupu Hospital, Shanghai (30 males and 30 females; mean age 10.67 +/- 2.47 years), and only right eyes were included in the analysis. The Swan 660 M was used to measure central axial length and peripheral axial length. Measurement variability across operator and time was evaluated using the intra-group SD, test-retest repeatability, coefficient of variation, and intraclass correlation coefficient. Results: The Swan 660 M demonstrated good repeatability and reproducibility for the measurements of central axial length and most peripheral axial length points, including Nasal 1, Temporal 1, Temporal 2, and Temporal 3 (coefficient of variation <5%, intraclass correlation coefficient >0.9). However, measurement reliability was lower for certain peripheral axial length points, such as Nasal 5 (intraclass correlation coefficient <0.75). The device also showed high stability in the measurements of central axial length (intraclass correlation coefficient = 0.999). Conclusions: The Swan 660 M demonstrates high repeatability and reproducibility for central axial length and most peripheral axial length measurements in children, suggesting its potential application in tracking the progression of myopia and investigating the relationships between peripheral retinal morphology and refractive status in this population.