This article investigates whether the alteration in the rules for calculating the COFINS tax has affected the profits of companies listed on the Sao Paulo Stock Exchange (BOVESPA), subdivided into 18 economic sectors. We have also studied the effects of this change on the revenue raised by COFINS and on overall federal tax revenues. The study has an explanatory nature and instrumental experimentation is theoretical and empirical, it makes use of regression analysis of time series and panel data with a confidence interval of 95% in the microeconomic level. In order to study macroeconomic we have used regression analysis based on time series data seasonally adjusted. The study is relevant because changes in tax rules can affect the transactions of sectors of the real economy and can be inefficient regarding the allocation of their additional costs. The statistical results show that for 7 of the 18 sectors investigated, the change in the COFINS calculation rules (by which the rate was increased but the tax was made non-cumulative, with offsetting debits and credits) had an effect on firms' profits. The statistical results also reveal there was a real increase in COFINS revenue and total federal tax revenue, indicating the economic inefficiency of this taxation model, unlike what has been stated by economic agents of the federal government. Key words: COFINS. Profits. Tax revenue.