Abstract: The economic growth of Suzhou foreign capital manufacturing industry is subjected to numerous influential factors,and among these factors there is a multi-related relaxionship between variables,generally independent variable of multivariate regression analysis is more serious in multicollinearity.This paper uses partial least square regression to set up a model,and analyics the economic growth impact factor.so on the one hand,it generalizes the data information with maximum limit in the original independent variable system,overcome multicollinearity;on the other hand  independent variable has the greatest interprefaction  ability for dependent variable,and reliability is high.

Key words: Suzhou foreign capital manufacturing, partial least square regression, modeling, factor analysis