Global change and forest succession are likely to change litter input to forest soils, including litter quality and litter amount, but the effects of such changes on CO_2 emission at forest floor still remain largely unknown. We initiated a manipulated experiment of litter in three successional subtropical forest (Masson pine forest (early stage), mixed forest (interim stage) and monsoon forest (advanced stage) in Dinghushan Nature Reserve. 5 replicates of litter remove, litter double, litter replacement (e.g., litter from mixed forest and monsoon forest take the place of original litter in Masson pine forest), and control treatments were established in each forest type. CO_2 emission at forest floor was measured using static chamber and gas chromatography techniques and air temperature, soil temperature and moisture were determined during the study period. We found that: (1) Litter remove decreased CO_2 emission at forest floor in each of three successional forests, while litter double increased CO_2 emission at forest floor in each of three successional forests. (2)The response of CO_2 emission at forest floor to litter replacement was very different in the three forests. Litter replacement from Masson pine forest or mixed forest in monsoon forest increased CO_2 emission at forest floor. Litter replacement from Masson pine forest or monsoon forest in mixed forest decreased CO_2 emission at forest floor. Litter replacement from monsoon forest in Masson pine forest increased CO_2 emission at forest floor. Litter replacement from mixed forest in Masson pine forest decreased CO_2 emission at forest floor. (e)Both litter quality and amount affected CO_2 emission at forest floor through the changes of soil temperature and moisture. (4) CO_2 emission at forest floor increase exponentially with soil temperature under all treatments in each forest. Different treatments can alter differently the sensitivity of CO_2 emission at forest floor to soil temperature.