In the composition of fuel prices three components are distinguished: raw material costs, distribution costs and taxes, and changes in any of them have a significant impact on prices. In this paper a tool to predict the final prices of fuel in the retail market is presented. The estimation is made from crude oil prices, raw material costs, taking into account historical data of price before taxes, including distribution costs and profit margins, and differentiating the geographical entity, region, province or municipality, which is to predict, allowing taxes to get into play.