The performance of the financial institutions is a major concern for both, the regulators and the policy makers, since it has a strong linkage with the performance of the economy. Banking has the same features like any mature industry. Companies that survive should keep their costs down. The banking sector is considered sensitive and important in any society, as it is the basement and the pillar of success or failure of any economy. Measuring efficiency of resources exploitation of this sector is considered necessary, to rationalize and direct the process of the banking industry for every country In this article we measure the bank efficiency based on the level of consistency between the head quarter and its branches from the perspective of achieving the main goals and sub goals of the main branch and its dependence respectively. The aim of this paper is to use a bi-level linear programming technique with rough parameters in the constraints to measure the productive efficiency of commercial banks. The proposed linear objective functions will be maximized for different goals that corresponding to the different banks. Based on D&K's goal programming method the described approach was developed to deal with the bi-level decision-making problem. The concept of tolerance membership function together with the branch and bound technique were used to generate the optimal solution for the problem under investigation. Also an auxiliary problem is discussed to illustrate the functionality of the proposed approach.