Identifying the primary regulators of plant litter decay in agricultural soils is important for understanding ecosystem function now and for managing post-harvest litter in a warmer world. We conducted a litter decay study at 11 sites in Canada and New Zealand with diverse soils and climates. We incorporated 13C-labelled barley litter into the surface 10 cm of soil and monitored amount remaining over 8–10 years; at five sites litter was also applied to the soil surface to simulate no-tillage. Our objectives were to evaluate litter decay as influenced by soil type, tillage practice, and environmental conditions, and quantify the most important factors controlling C retention in soil. Loss of C via decomposition occurred quickly – more than half was lost within 1 year and only about 5–12% remained at the end of the experiment. A double exponential decay model, based on temperature and developed from the first 5 years of the study, accurately described litter decay, but only after re-tuning using measurements from the full experimental period. Including precipitation in the model further improved its fit. Soil properties exerted minimal discernible influence on the amount of litter C remaining, implying that properties such as the amount and surface area of clay minerals, were less important than climate in limiting litter decay or enhancing retention of C in soil. Comparison of litter application treatments showed that no-till slowed decay of plant litter, but only for a short time (∼1 yr) and only in environments with a mean annual precipitation of <1000 mm. These findings have implications for the role of soils in climate mitigation. If only 10% of added plant litter C remains in soil beyond a few years, regardless of climate, residue placement, or soil type, then rates of soil gain are limited without substantive increases in residue inputs.