Purpose: This study investigated the trends of repetition and dropout rates in Myanmar's lower secondary education before and after the introduction of the "Continuous Assessment and Progression System (CAPS)" and probed the dependence of these tendencies on high-, middle- and low- socioeconomic status (SES). The obtained results were then examined to extract effective policy implications for the achievement of universal secondary education as specified in the Sustainable Development Goals. Design/methodology/approach: Before and after the CAPS introduction at four government secondary schools, grade repetition and dropout rate trends were examined with respect to differences in students' SES. The analysis utilised a sample of 7,272 students from target secondary schools in urban Yangon Region, Myanmar. Findings: It was found that since the introduction of CAPS, the grade repetition rates had fallen significantly in all SES groups, so was effective regardless of students' SES. The results also demonstrated the influence of unequal CAPS on dropout rates: in the middle-SES group, significant falls to nearly zero post-CAPS implementation. The high-SES group was at ceiling pre- and post-CAPS, so was unaffected. However, in the low-SES group, high dropout rates persisted, indicating that the poor socioeconomic backgrounds of these students significantly reduced the benefits of CAPS. Originality/value: Rather than using cross-sectional data such as education statistics, this study used longitudinal data based on academic enrollment registries that included information on individual enrollment statuses, which allowed for the relationships between grade repetition, school dropout, education policies and socioeconomic circumstances to be elucidated.