Abstract: This paper selects China’s 17 large-scale iron and steel enterprises as the research objects.On behalf of the capital structure, the debt ratio of total assets is to do the explained variable, using the growth rate of total assets, return on equity, total asset turnover and liquidity ratio to represent growth indicators, profitability indicators, operational indicators and debt indicators to do explanatory variables.The paper uses unit root and cointegration test stationary of sequence, using analysis of covariance to establish a fixed effects panel data model with variable coefficients, and then does descriptive statistics analysis for 5 variables of 17 enterprises, finally estimates and verifies the panel data model parameters, constructing the panel data model to estimate the relevant parameters, and studying the degree of influence of variable indicators to corporate capital structure.