The objective of this study is to investigate pollen-vegetation relationship in the Qilian Mountains. The eastern Qilian Mountains are located in the transitional zone of the Tibetan Plateau, the Loess Plateau and the arid region of Northwest China, which is one of the key areas of global environmental change. A total of 13 surface pollen samples from main vegetation have been collected. Pollen percentages were calculated in all samples. In order to reveal the relationship between pollen composition and the vegetation types from which the soil samples have been collected, Detrended Correspondence Analysis (DCA) ordination method was employed on the pollen data. The results show that dominating vegetation types can be recognized by their pollen spectra: Picea crassifolia forest, alpine shrub and alpine meadow as well. Altitude and temperature determine the distribution of the surface pollen and the vegetation. The good agreement between modern vegetation and surface samples across this area provides a measure of the reliability of using pollen data to reconstruct paleoenvironment and paleovegetation patterns in this or other similar regions. However the loss of Betula pollen in forest needs further investigation. Pollen oxidation is the most important factor contributing to the damage of modern pollen in the study area. Pollen concentrations decrease with the increase of pH values of soils, and decrease sharply when the pH exceeds 7.6.