Traditionally, company valuation methods are based on discounted cash flows, market prices, comparable sales and even liquidation values, but these are known to have a number of shortcomings. The application of data envelopment analysis (DEA) to finding companies comparable to the one under examination and there by predicting the market values of such companies can be considered to be an extension of the market-based approaches. As a result of using DEA, companies are classified into either an "inefficient" or "efficient" group, for each of which we assume that a corresponding upper or lower bound of its market value exists, respectively. Furthermore, the market value of an inefficient company is expressed in a form of a range of values, which are calculated by utilizing firm's reference set defined by DEA. As the validation to the modelling part, inefficient companies correctly classified their market values into the evaluated variation range up to 36.93%, and 66.20% of the actual market values are below their upper bounds. On the other hand, 41.67% of the time DEA can find alower bound of their market values for efficient companies based on inefficient peer comparisons. The results show that using DEA in valuing private companies is a relatively advanced method, and could prospectively play an active role in company valuation.