Abstract Conservation agriculture (CA) changes soil properties and processes compared to conventional agriculture. These changes can, in turn, affect the delivery of ecosystem services, including climate regulation through carbon sequestration and greenhouse gas emissions, and regulation and provision of water through soil physical, chemical and biological properties. Conservation agriculture can also affect the underlying biodiversity that supports many ecosystem services. In this overview, we summarize the current status of the science, the gaps in understanding, and highlight some research priorities for ecosystem services in conservational agriculture. The review is based on global literature but also addresses the potential and limitations of conservation agriculture for low productivity, smallholder farming systems, particularly in Sub Saharan Africa and South Asia. There is clear evidence that topsoil organic matter increases with conservation agriculture and with it other soil properties and processes that reduce erosion and runoff and increase water quality. The impacts on other ecosystem services are less clear. Only about half the 100+ studies comparing soil carbon sequestration with no-till and conventional tillage indicated increased sequestration with no till; this is despite continued claims that conservation agriculture sequesters soil carbon. The same can be said for other ecosystem services. Some studies report higher greenhouse gas emissions (nitrous oxide and methane) with conservation agriculture compared to conventional, while others find lower emissions. Soil moisture retention can be higher with conservation agriculture, resulting in higher and more stable yields during dry seasons but the amounts of residues and soil organic matter levels required to attain higher soil moisture content is not known. Biodiversity is higher in CA compared to conventional practices. In general, this higher diversity can be related to increased ecosystem services such as pest control or pollination but strong evidence of cause and effect or good estimates of magnitude of impact are few and these effects are not consistent. The delivery of ecosystem services with conservation agriculture will vary with the climate, soils and crop rotations but there is insufficient information to support a predictive understanding of where conservation agriculture results in better delivery of ecosystem services compared to conventional practices. Establishing a set of strategically located experimental sites that compare CA with conventional agriculture on a range of soil-climate types would facilitate establishing a predictive understanding of the relative controls of different factors (soil, climate, and management) on ES outcomes, and ultimately in assessing the feasibility of CA or CA practices in different sites and socioeconomic situations. The feasibility of conservation agriculture for recuperating degraded soils and increasing crop yields on low productivity, smallholder farming systems in the tropics and subtropics is discussed. It is clear that the biggest obstacle to improving soils and other ES through conservation agriculture in these situations is the lack of residues produced and the competition for alternate, higher value use of residues. This limitation, as well as others, point to a phased approach to promoting conservation agriculture in these regions and careful consideration of the feasibility of conservation agriculture based on evidence in different agroecological and socioeconomic conditions.