With the huge concerns on the inefficiency of public enterprises, particularly a significant amount of debt, an increasing number of studies have been carried out to analyze the levels of inefficiency and investigate the causes of that inefficiency. However, very limited range of analytical methodologies have been used in the efficiency analysis and moreover, the effects of external factors have been little addressed. This study explores the efficiency of urban railway corporations in Korea by utilizing a method of stochastic frontier analysis (SFA). In particular, the potential effects of external factors including residential and floating populations of a station were statistically analyzed. A total of seven Korean urban railway corporations were selected to compare each other in terms of operational efficiency. The results present three important findings. First, the Cobb-Douglas model was found to be more valid for SFA compared to the Translog model. Second, the efficiencies of urban railway corporations in Seoul and Busan are relatively high whereas those of Daejeon and Gwangju are very low in efficiency in the area of sales revenue. In an aspect of number of transport of passengers, Gwangju Metro also showed the lowest efficiency. Third, the external factors are significantly associated with the efficiency, indicating that the efficiencies of Daejeon Metro and Gwangju Metro would increase while the efficiency of Seoul Metro would decreases when the external variables are excluded in the efficiency analysis. The results provide several meaningful implications for managers of the urban railway corporations as well as policy makers who are attempting to resolve the inefficiency problems of public enterprises.