This study attempts to obtain efficiency scores of thirty-one railway companies operating worldwide by using data envelopment analysis. To assess the data covering a term from 2000 to 2009, we use CCR and BCC methods. According to the results of the CCR model analysis, 17 firms are found to have technical efficiency in the first year, whereas this figure goes up to 18 companies in the last year. Input oriented and variable return analysis in the concept of the BCC model, the number of the firms found technically efficient at the beginning of the period are 20. At the end of the period, this figure is up to 24. We implement panel regression analysis to estimate the effects of the companies' output on their efficiency. Results imply that CCR models provide us more meaningful explanations. Malmquist Index analysis also indicates that total factor productivity increases by 0.03 % for the entire period.