The Tiantong national forest park of Zhejiang Province boasts evergreen broad-leaved forest (EBLF) representative of the southeastern seaboard of China. Castanopsis fargesii is the most extensively distributed EBLF species in the subtropical zone and the main dominant species of the Tiantong forest. Five plots were established, each of which had a unique disturbance regime, to study photosynthetic ecophysiology of C. fargesii. The LI-6400 portable photosynthesis system was used. By comparative analysis of daily variances of net photosynthesis (Pn), transpiration (E), water use efficiency (WUE), and apparent quanta efficiency (AQE), we found that net photosynthesis and transpiration were higher in the tree-less plots (Plots 1, 2, & 3) than in the shrub-less plot (Plot 4) and the control plot (Plot 5). This might be a result of the high levels of photosynthetic active radiation (PAR) in the tree-less plots. WUE and AQE in plot 5 and plot 4 were higher than in the tree-less plots. Using Spss for windows (Version 11.5), we established correlation matrices of photosynthetic rate and factors affecting C. fargesii seedlings and saplings in the differently disturbed plots. Ta (ecological factor) and Gs (physiological factor) are the foremost factors affecting C. fargesii seedlings; PARi, Tair, Vpdl, and Gs are the main factors affecting C. fargesii saplings. Using the stepwise regression method, we established linear regression equations of photosynthetic rate and various affecting factors. At α=0.05, the two following equations are statistically significant and have good predictive power. C. fargesii seedings: Pn= -2.399 + 0.205Ta + 43.209Gs; C. fargesii saplings: Pn= -2.265 + 0.006PAR - 3.915Vpdl + 0.364Ta + 38.121Gs.