Fossil pollen records from 45 sites across China were evaluated and synthesized to document vegetation and climate change during Marine Isotope Stage 3(MIS 3)at 60–27 ka and to understand the large-scale controls.During MIS3, vegetation type was dominantly forest in eastern China, forest steppe/meadow in the north and Tibetan Plateau, and steppe desert in northwestern arid China. We developed a semi-quantitative vegetation index to reflect change in plant abundance(by inferring the general climate conditions), with a vegetation score from 1to 3 based on the different vegetation types inferred from pollen data at individual sites at intervals of 2,000 years.The reconstructed vegetation index shows higher values during MIS 3, especially during the period 53–40 ka, than at the Last Glacial Maximum. Our results also suggest that climate on the basis of vegetation change was cooler and drier during MIS 3 than during the Holocene optimum;however, MIS 3 vegetation was probably similar to modern vegetation. The close relationship between vegetation change, insolation and Asian summer monsoon strength suggests that climate variations, probably in both temperature and precipitation, are the primary drivers of regional vegetation change. Additional well-dated, high-resolution palaeoclimate records from many locations across China will be needed to understand the vegetation change and climate forcings on millennial and centennial scales within MIS 3. Fossil pollen records from 45 sites across China were evaluated and synthesized to document vegetation and climate change during Marine Isotope Stage 3 (MIS 3) at 60-27 ka and to understand the large-scale controls. Fluid MIS3, vegetation type was dominantly forest in eastern China, forest steppe / meadow in the north and Tibetan Plateau, and steppe desert in northwestern arid China. We developed a semi-quantitative vegetation index to reflect change in plant abundances (by inferring the general climate conditions), with a vegetation score from 1to 3 based on the different vegetation types inferred from pollen data at individual sites at intervals of 2,000 years. Reconstructed vegetation index shows higher values ​​during MIS 3, especially during the period 53-40 ka, than at the last Glacial Maximum. Our results also suggest that climate on the basis of vegetation change was cooler and drier during MIS 3 than during the Holocene optimum; however, MIS 3 vegetation was probably similar to modern vegetation. The close relationship between vegetation change, insolation and Asian summer monsoon strength suggests that climate variations, probably in both temperature and precipitation, are the primary drivers of regional vegetation change. Additional well-dated, high-resolution palaeoclimate records from many locations across China will be needed needed to understand the vegetation change and climate for cings on millennial and centennial scales within MIS 3.