The study on soil texture space variation is key to regional soil moisture solute cycle simulation. Condition simulation built on a Markov chain is a geostationary simulation method using multiple information sources, and has several advantages compared to the traditional interpolation and variogram-based conditional simulation. In this study, 1 020 soil samples from 102 soil profiles in FengQiu County were collected, and 3D spatial distribution models of soil texture types were built by using the transition probability method. On the basis of 500 realizations of stochastic simulation, the maps of soil texture type space distributions and their probability distributions were got. The results showed that 1D dimensional vertical transition probability was an effective way to describe the asymmetry and anisotropy of various soil spatial distributions in perpendicular direction of the study area. Using the approach of transition probability to describe the uncertainty in the spatial distribution of soil texture type could effectively overcome the shortcomings of traditional spatial interpolation.