Dryland area accounts for approximately 40% of worldwide land area, which plays a significant role in regulating the carbon sequestration capacity of land. Vegetation restoration in drylands adopted to prevent land degradation, and may also serve as a carbon sink in the earlier stage. However, the persistence of the carbon sink for the revegetated ecosystem in drylands is still unknown. Can the well-established restoration vegetation in dryland areas serve as a carbon sink in long-run? To address this question, we investigated the carbon sequestration capacity of planted vegetation in dryland areas with 13 years of observation (2009–2021) for established vegetation restoration, which began in 1989. We found that the revegetation area serves as a carbon sink in all years. The mean annual net ecosystem productivity (NEP) is 91.61 ± 36.17 gC m−2 yr−1 (mean ± standard deviation). Annual NEP showed a significant increasing trend over the study period with a rate of 5.65 gC m−2 yr−1 yr−1 (p<0.05). The increase in spring temperature, the earlier start of net carbon uptake, and the longer duration of net carbon uptake contribute to the gradual trend of NEP. The amount of annual NEP is predominantly determined by summer precipitation. Meanwhile, our results revealed that the increase in net carbon uptake by revegetation did not lead to excessive consumption of water resources. Our results have suggested that appropriate vegetation restoration in arid areas can increase ecosystem carbon sequestration over longer timescales and mitigate climate change, with relatively low environmental consequences and risks. Considering the vast area of degraded land in the global drylands, the carbon sequestration effect of this model should be given more attention .