In order to test the accuracy of the usually-used fixed calibration factor of the canopy scanner of LAI-2000 for measuring the leaf area index (LAI), a Larix principis-rupprechtii plantation was chosen in the small watershed of Xiangshuihe located at the Liupan Mountains of Ningxia Hui Autonomous Region of NW China, the LAI was measured in October 2010, a period from full canopy to the total fall of needles, by using both the LAI-2000 and litter-fall collection method. Then, a comparison was made between the LAI values determined by the litter-fall collection and that calculated based on the figures read from LAI-2000 and the fixed calibration factor (1.49). It showed that the average of LAI measurements of the 2 methods was very close, with a difference of only 5%. However, the calculated LAI from LAI-2000 was obviously higher than the true values determined by litter-fall collection when the canopy was full of needles; and obviously lower than the true value when the canopy was sparse after needle falling. The reason may be that LAI-2000 takes the projection of twigs as needles. So, a dynamic calibration factor is needed, especially in the seasons when the needle amount and the percentage of twigs projection in crown projection change quickly. Therefore, a statistic relation in a quadratic polynomial form between the 2 series of LAI data was well fitted. This relation can be used for a more accurate estimation of LAI based on the data read from the easily-used canopy scanners like LAI-2000.