Automated Teller Machines (ATMs) are one of the most important key touch-points to reach the customers. The investment in ATMs and the impact on the banking industry is growing steadily. Therefore, in order to increase the profitability, banks seek to make their investments in the right places for ATMs. Banks have to take into consideration many factors such as traffic flow, competition, cost and safety in order to determine the optimum locations of ATMs. This paper proposes a decision support system that determines the importance value of ATM deployment decision criteria which can be used to decide the best ATM locations afterwards. Fuzzy Analytic Network Process (FANP) methodology is preferred to evaluate the importance levels of deployment criteria and Chang's extent analysis is employed for eliminating the vagueness of the decisions from the evaluation process. The decision-makers are asked to express their opinions on the comparative importance of various factors in linguistic terms. These linguistic variable scales are then converted into fuzzy numbers, since it becomes more meaningful to quantify a subjective measurement into a range rather than an exact value. Finally, the relative importance of the criteria is obtained. The decision process includes the identification of 17 criteria grouped into 5 clusters which are named financial, commercial, traffic, demographic and strategic. The application of the proposed method was performed for Turkish banking industry. Number of Bank's ATMs, Expected Level of Commission Income and Expected Level of Transaction Income were found as the three most significant criteria that should be analyzed deeply by the decision makers in ATM location selection decision. The results of the proposed method considered applicable and valid by the experts who included in the decision process and this analysis demonstrated by their responses to a set of questions.