warming associated with a huge input of reduced carbon into the ocean, atmosphere and biosphere. The magnitude of carbon release during the PETM was very similar to the present anthropogenic carbon dioxide emission by burning fossil fuels. Current data estimate that the total amount of carbon released during the PETM was similar in magnitude to the IPCC RCP8.5 emission scenario. Therefore, a comprehensive study of the PETM provides a unique opportunity for understanding the relationship between carbon emission and climate change, and may help to quantitatively assess the carbon sequestration potential and rate of changes in the ecosystem, ocean and lithosphere. So far, the basic outline of massive carbon release and extreme warmth during the PETM is well established, but many important aspects regarding the carbon cycle during the event remain unresolved. The source of the carbon is still debated, with hypothesized carbon sources including methane hydrates in the ocean floor, soil organic carbon in circum-Arctic and Antarctic terrestrial permafrost, shallow-buried peat and coal, volcanic degassing, thermogenic methane caused by volcanic intrusions, and a cometary impact. The controversy can be encapsulated by two questions: (1) Is the PETM unique or only one case among many similar events? (2) What was the coupled relationship between the fine structure of the carbon isotope excursion and environmental change during the PETM? Uncertainty about the source also creates uncertainty about the amount and rate of carbon released, and the fate of the massive excess carbon at the end of the event is also poorly constrained. Here, we summarize and analyze carbon-cycle dynamics during the PETM and draw three preliminary conclusions. First, a massive release of C-12-enriched carbon operated as a positive feedback to the temperature increase during the PETM, implying that methane hydrates and/or permafrost were plausible carbon sources. In particular, peak temperatures of short duration are associated with minimum carbon isotope values, suggesting that the carbon source was probably exhausted, and the positive feedback ceased. Second, the total amount of carbon released during the PETM was in the range of 4000-10000 Gt C, reflecting considerable uncertainty caused by the large range of the magnitude of the carbon isotope excursion. The average rate of carbon emission during the PETM is estimated to have been in the range of 0.15-1 Gt C/a, an order of magnitude lower than the present rate of anthropogenic carbon emission, implying that current human activities may trigger a positive feedback of the Earth's surface system and intensify global warming. Therefore, to avoid the initiation of the potential feedback of the warming-induced decomposition of reduced carbon reservoirs, as well as the accompanying major environmental perturbation, it is urgently necessary to reduce or slow the carbon emissions. Third, excess C-12-enriched carbon was removed from the exogenic carbon pool during the PETM recovery on a timescale of similar to 40 ka, much faster than the similar to 100 ka that is expected for carbon uptake only by silicate weathering. These observations suggest that the fertilization effect of CO2 on plant growth, the regeneration of carbon stores in the high-latitude terrestrial biosphere, and the enhanced efficiency of the marine biological pump accelerated the rate of carbon sequestration and the subsequent recovery from the PETM.