During the past 40 years, more than 400 000 hm² of the tide flat in the Yangtze River delta of China have been reclaimed for agriculture. The polders must be washed to have the salt in the soil leached away before they can be used to grow rice. However, strong evapotranspiration brings salt back from deep soil layers, and the salt accumulates in the root-zone (0~1 m). So accurate three-dimensional soil salt maps can serve as decision-making basis for rice farmers of the polders to work out rational cultivation programs. So far little has been reported on three-dimensional space variability of soil salts and most of the researches in this field at home and abroad are using two-dimensional space variation maps to express three-dimensional space variability. How to express three-dimensional space variability of soil properties is a challenge to the traditional soil profile sampling, space variation analysis and three-dimensional visualized expression. In this project, the paddy field in the polder of Zhejiang Province was cited as a case of study. First, apparent electrical conductivity (ECa) of the paddy field was measured using an EM38 from different heights above the ground and then a linear model coupled with the Tikhonov regularization method was used to inverse electrical conductivities at 10 different soil depths of a profile of 0~110 cm. The data gathered from the 56 soil profiles in the field were used as basis in the study on three-dimensional space variability of soil salinity. Then, the soil ECa was interpolated three-dimensionally on a field scale using the Three-dimensional Inverse Distance Weighting (3D-IDW) method. In the end, the sphere, section and plume models of the Virtual Reality Modeling Language (VRML) were used separately to build soil electrical conductivity virtual reality visualized models for scattered electrical conductivity points on profiles, sections of two dimensional space variation at the horizontal and vertical directions and three-dimensional soil solum model, and eventually to realize distribution of VRML models on the internet (http://agri.zju.edu.cn/3d/). The descriptive statistics of profile soil salinity shows that the minimum, median and maximum values of EC all increased with soil depth. On the field scale, 3D-IDW is a fairly good method for prediction and interpolation of three-dimensional space distribution of soil EC, and VRML is a good one, too, for visualized modeling to exhibit and analyze three-dimensional space distribution laws of soil EC. Horizontally, soil salinity increased gradually form the northwest to the southeast of the filed, and vertically, it did with soil depth. Soil salinity was the highest in the southeast corner of the field. The spatial distribution of soil salinity is true, The drainage ditches around of the field helped drain soil water from the field, lowering the soil humidity. Besides, the field declined from northwest to southeast in topography. River water was led into the field in the west to wash soil salt and drained through the trunk into the Hangzhou Gulf. As the northwest corner was higher in topography, it got free of the impact of sea water earlier and soil salt therein was easily leached off, while salt-containing groundwater flew continuously toward the lower southeast corner. Users can access the website for VRML models, and do some basic operations, such as pan, zoom in, zoom out and rotation, to the VRML models. With the aid of EM38, 3D-IDW and VRML it is feasible to plot visualized three-dimensional soil properties space variability maps, which can be shared via network.