Management of terrestrial carbon fluxes is being proposed as a means of increasing the amount of carbon sequestered in the terrestrial biosphere. This approach is generally viewed only as an interim strategy for the coming decades while other longer-term strategies are developed and implemented — the most important being the direct reduction of carbon emissions. We are concerned that the potential for rapid, disturbance-induced losses may be much greater than is currently appreciated, especially by the decision-making community. Here we wish to: (1) highlight the complex and threshold-like nature of disturbances — such as fire and drought, as well as the erosion associated with each — that could lead to carbon losses; (2) note the global extent of ecosystems that are at risk of such disturbance-induced carbon losses; and (3) call for increased consideration of and research on the mechanisms by which large, rapid disturbance-induced losses of terrestrial carbon could occur. Our lack of ability as a scientific community to predict such ecosystem dynamics is precluding the effective consideration of these processes into strategies and policies related to carbon management and sequestration. Consequently, scientists need to do more to improve quantification of these potential losses and to integrate them into sound, sustainable policy options.