[Objective] To study the characteristics of soil seed bank(SSB) in Baxian Mountain in Jingxian County of Tianjin City in order to provide reference for vegetation restoration using SSB.[Methods] As a new correspondence analysis method, the detrended canonical correspondence analysis(DCCA) method can overcome the arch effect and give better explanation between plant communities and environmental factors. [Results] (1) Compared with other plots of Tianjin City, the species richness of SSB was very high, reaching to 42 species, but the density of SSB was the lowest. The bare rock, thin soil and obvious animal feeding effect were main reasons for the low density. (2) The Sorensen similarity index was very low, but the Shannon-Wiener diversity index, the Simpson index and the Pielow index were very high. (3) With the increase of the altitude, the community structure changed rapidly, the number of woody plants increased. With the slope exposing to the sun, the type of plants changed from shade plants to light plants.[Conclusion] By DCCA analysis, three obvious clusterings were presented. Organic matter, pH value and total phosphorus were main soil factors which affected the distribution of SSB.