This paper estimates the Marginal Abatement Cost Curve (MACC) of CO₂ emissions in China using provincial panel data from 2001 to 2010. The study uses a directional output distance function to estimate the provincial marginal abatement cost (MAC) of CO₂ emissions. Four types of model specifications are applied to fit the MAC-carbon intensity pairs, and the optimal specification is identified econometrically. A scenario simulation of China's 40-45% carbon intensity reduction based on the MACC is illustrated. The simulation results show that China would incur a 559-623 Yuan/ton (roughly 51-57%) increase in marginal abatement cost to achieve a corresponding 40-45% reduction in carbon intensity compared to its 2005 level. The paper compares four types of MACC specifications and selects the optimal one based on in-sample and out-of-sample performance. The study finds that the quadratic functional form performs best in terms of in-sample fitness, while the logarithmic form is optimal for out-of-sample forecasting. The results indicate that the marginal abatement cost curve is downward sloping and convex, meaning that the cost of reducing additional units of CO₂ emissions increases as carbon intensity decreases. The study also simulates the cost of achieving China's carbon reduction target of reducing carbon intensity by 40-45% by 2020 compared to 2005 levels. The simulation results show that the marginal abatement cost would increase significantly under different scenarios, with the BAU scenario showing the highest increase. The results suggest that China would need to invest heavily in abatement measures to achieve its carbon reduction targets. The study concludes that the quadratic functional form is the optimal specification for policy analysis, as it provides a more flexible and accurate representation of the marginal abatement cost curve.This paper estimates the Marginal Abatement Cost Curve (MACC) of CO₂ emissions in China using provincial panel data from 2001 to 2010. The study uses a directional output distance function to estimate the provincial marginal abatement cost (MAC) of CO₂ emissions. Four types of model specifications are applied to fit the MAC-carbon intensity pairs, and the optimal specification is identified econometrically. A scenario simulation of China's 40-45% carbon intensity reduction based on the MACC is illustrated. The simulation results show that China would incur a 559-623 Yuan/ton (roughly 51-57%) increase in marginal abatement cost to achieve a corresponding 40-45% reduction in carbon intensity compared to its 2005 level. The paper compares four types of MACC specifications and selects the optimal one based on in-sample and out-of-sample performance. The study finds that the quadratic functional form performs best in terms of in-sample fitness, while the logarithmic form is optimal for out-of-sample forecasting. The results indicate that the marginal abatement cost curve is downward sloping and convex, meaning that the cost of reducing additional units of CO₂ emissions increases as carbon intensity decreases. The study also simulates the cost of achieving China's carbon reduction target of reducing carbon intensity by 40-45% by 2020 compared to 2005 levels. The simulation results show that the marginal abatement cost would increase significantly under different scenarios, with the BAU scenario showing the highest increase. The results suggest that China would need to invest heavily in abatement measures to achieve its carbon reduction targets. The study concludes that the quadratic functional form is the optimal specification for policy analysis, as it provides a more flexible and accurate representation of the marginal abatement cost curve.