The main objective of this work, based on a quantitative analysis, is to show the relationship between the return on the shares of Brazilian state-owned companies (Banco do Brasil, Eletrobras and Petrobras) and macroeconomic and economic-financial variables, as well as to quantify the impact that these variables have on the return on shares, in the period from August 1994 to July 2014. To carry out this research, multiple linear regression was used, with the aid of the correlation test and the Variance Inflation Factors (VIF) test for collinearity analysis. The multiple linear regression models showed that some of the variables had very high p-values, so some variables were excluded so that the model had a high level of confidence. However, the existence of influential variables can be observed both in the study with macroeconomic variables and in the study with economic-financial variables, where the lowest coefficient of determination was just over 50% and the largest model with the highest coefficient was 99%.