Land use is the linkage between human and nature.It is subject to natural environment and is seriously influenced by social and economic activities.So the study on land use patterns and their influential factors has become indispensable in LUCC research.Heavy interaction of research data and inadequate observation samples are the key problems to this study.Thus it is very important to optimize the research methods. Partial Least Squares (PLS) regression is a novel multivariate data analysis method developed from practical applications in real world.It is mainly used to model linear regression between multi dependent variables and multi independent variables.PLS regression has many advantages, which traditional methods do not have,such as avoiding the harmful effects in modeling due to the multicollinearity and regressing when the number of observations is less than the number of variables,etc.Moreover,it can combine the basic functions of regressing model,principal compo nents analysis (PCA) and canonical correlation analysis.Therefore,PLS regression is a optimized tool to LUCC research. In this paper,PLS regression is applied to study the relationship between land use patterns and their influential factors,taking Suxichang region as a case study.Supported by SAS/STAT statistical software,the PLS regression model was built and the calculated results were meticulously analyzed.Though the accuracy analysis of PLS factors,the most important PLS factors were selected,used in studying the relationship between each type of land use pattern and its influential factors.In addition,a predict model about land use pattern change was produced and analyzed. This case study confirmed the practicability and operability of the PLS regression method.It showed that PLS regression has the ability to make quantitive research on land use patterns,and could be widely applied to geographical research.