Conservation tillage of farmlands is very important to soil building, guarantee of food security and mitigation of global warming. A total of 303 sets of field experiment data was collected from 157 long-term (1980 – 2012) conservation tillage experiment sites for meta-analysis to characterize quantitatively changes of SOC (soil organic carbon) in the topsoil (0~15cm for paddy and 0~20cm for upland) of farmlands under conservation tillage. Results show that the practices of conventional tillage plus straw incorporation (CTS), no tillage (NT), and no tillage plus straw incorporation (NTS) significantly increased SOC content in the topsoil with a relative change rate (RC) being 0.22, 0.35, and 0.52 g kg^-1 a^-1, respectively, as compared with the practice of conventional tillage (CT). The effect was more significant in paddy fields than in uplands and in farmlands under the double cropping system than in those under the single cropping system. It seems, however, that SOC accumulation and its amplitude is not fully related to the initial content of organic carbon in the soil. The SOC increase rate in short-term field experiments (≤5a) is about 1.75 times as high as that in long-term ones (>5a). It is, therefore, possible to overestimate SOC sequestrating potential of the conservation tillage practices just based on short-term field experiments.