Spatial variability of soil moisture at different layers in the coastal region of north Jiangsu Province was studied by using geostatistics in this paper. Cross-semivariogram was then used to analyze the coregionalized relationship between surface soil moisture and moisture at deep soil layers, and estimation of soil moisture by Kriging and CoKriging was discussed. Results indicated that soil moisture exhibited weak variation and moderate spatial autocorrelation owing to compound impact of structural and stochastic factors. Coregionalized variables both showed positive correlation and the spatial structure of coregionalized variables was superior to that of univariable. Short-range variability of soil moisture was depicted more detailedly by CoKriging than ordinary Kriging. Compared with ordinary Kriging method, the root-mean-square error produced by CoKriging decreased by 10.1% to 21.3%, the average standard error decreased by 11.3% to 17.7%, and the determination coefficient between the observed values and the predicted values increased by 15.9% to 26.4%. The research results can serve as a theoretical reference to the regional management of soil moisture and optimal regulation of soil water and salt in coastal region.