According to a recent study by GfK Romania, the institute number one of market research in Romania, 55% of Romanians that plan to buy a car and would buy a second hand car. Given the high demand for such cars, this study is to facilitate the acquisition of second hand cars. This will be achieved through analysis of existing data on the auto market using data mining techniques. The paper has determined the price of a car using linear regression and the of score each type of car, which, on a scale from 1 to 5, will show whether or not a car worths to be bought. Also there have been identified the most important criteria considered when choosing a car using the algorithm "InfoGainAttributeEval" Ranker search method. The results obtained indicate a strong correlation between price and features set cars: class, fabrication_year, no_km, combustible, engine capacity, registration_statementand and emissions_class.