Agricultural carbon (C) pool is one of the most active parts in global terrestrial ecosystem C pool and is very sensitive to change in management measures. As an important means for dry farming in China to get high yield, irrigation and its change in patterns certainly will have a significant impact on the ecosystem C budget and its atmospheric greenhouse effect, but the relative data and corresponding mechanism analysis are rarely known. This paper reviews research progresses in the quantitative impacts of irrigation on soil respiration and its different components, the fluxes of nitrous oxide and methane as well as soil C storage, and compares the differential responses of the above-mentioned processes to different irrigation patterns, amounts, frequencies, lasting periods and irrigation water quality, etc. The results indicate that irrigation generally will increase soil C emission, but the relationship between irrigation quota and soil carbon dioxide (CO_2) effluxes is nonlinear. Water additions stimulate soil CO_2 emission when soil is under water deficit condition, whereas excessive irrigation would decrease soil respiration rates. Irrigation increases the sensitivity of soil respiration to temperature change (Q_(10)), and the response of Q_(10) to the irrigation amounts can describe as a parabola. The Q_(10) values under drip irrigation are larger than those under flood irrigation. There is a great difference in the response of different respiration components to the change of soil water under different irrigation methods, and the effect of irrigation on root respiration is more significant than on soil microbial respiration. The ratios of root respiration contribution to soil respiration under drip irrigation are much larger than those under flood irrigation. Irrigation water quality and irrigation depth also have an important impact on total emission amount of greenhouse gases, and full municipal wastewater treatment can obtain much more carbon emission credits for countries and regions. Water management is one of the important measures that mitigate the emissions of nitrous oxide (N_2O) and methane (CH_4). Both the emission amount and the emission pathway of N_2O and CH_4 are significantly affected by the soil water condition. There usually exists marked trade-off relationship between CH_4 effluxes and N_2O effluxes when soil moisture changes, so the comprehensive evaluation indicators such as global warming potentials (GWPs) are more conducive to reflect the actual change in greenhouse effect contribution of agricultural soil exactly. There exist various possibilities in the effect of irrigation on soil organic C (SOC) pool, such as increase, decrease, or no significant changes, and the responses of SOC to irrigation also vary a lot under different climate and soil conditions. In general, the stimulated effects of irrigation on SOC are more significant in relatively dry regions than in humid regions. Meanwhile, SOC in different forms often shows different response sensitivities and variation trends to irrigation. The effect evaluation of irrigation measures should be considered from the multiple perspectives of water-saving, increasing SOC pool and enhancing the utilization efficiency of liable organic C. To sum up, there still exists large uncertainties in the effects of irrigation on soil greenhouse gas emission and SOC in agro-ecosystem up to now. The future study should be focused on following contents: (1) to strengthen the comparison study of different irrigation patterns, especially the study on the spatial differences of greenhouse gases emission under different irrigation patterns; (2) to pay more attention to the comprehensive effect of different greenhouse gases and the ecosystem carbon budget; (3) to develop in-situ field research for both long period and short period; (4) to probe into the microbiology driving mechanism of irrigation on farmland greenhouse effect and soil C sequestration.