Plants are the main source of soil organic matter in all ecosystems and are thus the main potential lever to increase soil organic matter stocks in agricultural soils in order to improve soil fertility and productivity, adapt to climate change and mitigate climate change. The effect of main cropped species on soil organic carbon (SOC) stocks has been described and incorporated in models for decades and the consequences of returning or not crop residues to soil is widely documented. Beyond this, a range of results show that there is still un-used potential in the management of plant inputs to agricultural soils in order to increase soil organic matter and SOC stocks, and that knowledge gaps are associated. The magnitude of SOC stocks results from the balance between OC inputs to soil and OC outputs by mineralization or erosion. Several experimental results and meta-analyses show that increases in SOC stocks with changes in management are more related to increased plant inputs than to decreased outputs by mineralization due to no-tillage. It is well established now that the yield in soil carbon is greater for below-ground OC plant inputs than for above-ground, showing that SOC oriented management should focus on the below-ground compartment. Increasing C belowground inputs to soil could be achieved through cropping deep rooting crop varieties, deep-rooting perennials, and agroforestry. For this, the balance between increased OC inputs at depth by root systems and consequent priming effect needs to be established and the carbon storage potential of sub-soils should be quantified. More generally quantifying rhizodeposition, investigating the role of root systems architecture and that of mycorrhizae in SOC storage and stabilization warrant further research. SOC storage is increased by plant species diversity and by the presence of legumes in the rotation, while these effects need to be confirmed, and the processes behind to be unraveled. Indirect effect of plant OC inputs on SOC storage, by affecting the physiology of soil microorganisms and in particular their carbon use efficiency also need attention. Managing soil organic matter for better SOC sequestration in agriculture via plant inputs hence involves increasing the crop biomass production and returns to soil as known for long, but also managing complex plant associations in space and time (cover crops, trees, hedges, associated crops..) and better explore the whole soil profile.