[Objective] To systematically summarize the existing sampling methods, and to explore an optimized sampling strategy to balance the sampling intensity, analysis cost and research precision, which was expected to maximize the accuracy of experiments with minimum economic input.[Methods] On the basis of extensive review on the domestic and foreign literatures in recent years, the optimized sampling strategy for soil properties was summarized. The theory of optimal sampling strategy was elucidated in two parts:the reasonable sampling number and the sampling point layout. Three methods and four modes were introduced to determine the reasonable sampling number and sampling points, respectively, and the future research was prospected.[Results] Most of the sampling program design used in current methods was model-based, and of which the spatial correlation of samples was ignored and relevant information was unused, these all were responsible to the deviation of experimental results and wastes of labor and materials to some extent. In addition, most researches were county scale or and/or lower scale, and there was no unified evaluation system or standards to design a formal sampling program.[Conclusion] In the future, the priority for optimal sampling design should be strengthened.