The Big Spring Number Eight (BSNE) sampler is an important and widely used instrument in studies of aeolian sand transport. Its sampling efficiency is an important parameter and varies with particle size. However, little is known about the variations in sampling efficiency caused by the differences in particle size with height. The variations in particle size with height from 10 to 200 cm were obtained from a field experiment at four sites in the Taklimakan Desert from May 2010 to April 2011. The variation in the sampling efficiency of the BSNE sampler with particle size was estimated based on previous studies and a wind tunnel experiment. It was found that the average particle size decreased with height above ground from 10 to 200 cm at three experimental sites, while at the other site it decreased from 10 to 15 cm and then slightly increased. The percentage distribution of particle sizes with height was similar for sizes below 63 μm and between 125 and 200 μm at all experimental sites, but there were differences in the percentage of size classes between 63 and 125 μm and above 200 μm. The sampling efficiencies ranged from 66.27 to 58.98%, 64.97 to 57.47%, 56.04 to 52.76%, and 72.07 to 58.9% at the four sites, respectively, with an average of 64.84 to 57.03% at heights from 10 to 200 cm. After correcting for sampling efficiency, the amount of aeolian sand that passed through a section of 100 cm (width) × 200 cm (height) increased by 44.5, 52.1, 78.8, and 26.4% at the four sites, respectively, with an average of 50.5%. These results indicate that the sampling efficiency of the device needs to be considered when making observations across a gradient of particle sizes.