Supplementary private tutoring (PT) is commonly implemented worldwide, particularly among countries with highly competitive school systems and high-stakes examinations. Families often consider PT to promote their children's skills and address any gaps in children's learning. However, PT generally requires a considerable amount of family resources, and whether PT effectively increases students' academic achievement remains subject of debate. This study examined student-report questionnaires and school reports of mathematics achievement from approximately 1500 secondary school students in China to track retrospective longitudinal variations of PT over a year. The study involved an analysis of the differential effectiveness of regular mathematics PT participation on the mathematics achievement of students from different quantiles at four hierarchical cognitive levels. The results indicated that only the lower-quantile students benefited from PT for their lower cognitive level achievement. Further, it was observed that among the higher-quantile students there was a reduction in high cognitive level achievement. These results raise questions concerning the role and effectiveness of PT. This study provides an evidence-based assessment of the effectiveness of PT and highlights the implications for educational practice and future research.