Several crucial acceleration periods exist during aging process. Epigenetic clocks, serving as indicators of aging, are influenced by genetic factors. Investigating how the genetic contributions on these clocks change with age may provide novel insights into the aging process. In this study, based on 1084 adult twins from the Chinese National Twin Registry (CNTR), we established structural equation models (SEMs) to evaluate the trends in genetic influence with aging for epigenetic clocks, which include PC-Horvath, PC-Hannum, PC-PhenoAge, PC-GrimAge, and DunedinPACE. A decline in overall heritability was observed for all five clocks from ages 31 to 70, with a relatively stable trend at first. Subsequently, apart from PC-GrimAge, the other four clocks displayed a more evident drop in heritability: DunedinPACE and PC-PhenoAge experienced a clear decline between 55 and 65 years, while PC-Horvath and PC-Hannum showed a similar decrease between 60 and 70 years. In contrast, the heritability of PC-GrimAge remained stable throughout. An analysis of methylation sites (CpGs) from these clocks identified 41, 26, 4, and 36 CpG sites potentially underlying heritability changes in DunedinPACE, PC-Horvath, PC-Hannum, and PC-PhenoAge, respectively. Data from the CNTR were collected through two surveys in 2013 and 2018. Based on 308 twins with longitudinal data, declines in genetic components were observed at follow-up compared to baseline, with significant decreases in the four PC-clocks. DunedinPACE peaked in 5-year longitudinal genetic contribution changes at age 55-60, while PC-clocks consistently peaked at age 50-55. These findings may offer novel insights into the role of genetic variations in aging.