Abstract: Automated Teller Machine (ATM) is the most important banking technology than ever before in order to access the bank account at anytime anywhere. Nowadays, the bank customer used telecommunication devices to make money transfer and transactions by touching few buttons on the screen without needs of visiting bank premises and cashier. Many financial institutions including Commercial Bank of Ethiopia (CBE) were now incorporating ATM and financial services as a key component of their growth strategy. However due to increase in the number of intelligent criminals and fraudsters, the ATM system was insecure. ATM systems today used no more than an access card and Personal Identification Number (PIN) for identity verification in CBE. Under this system sometimes there were security challenges like card and currency fraud. Thus, more reliable ATM system model in providing security by integrating facial recognition technology was proposed in order to solve insecurity. The developed system model protected customers and financial institutions alike from intruders and identity thieves. This thesis proposed an ATM security system model that was combined a physical access card, a PIN, and electronic facial recognition. Generally, to accomplish this security system model for CBE ATM system; the MATLAB program was developed. Keywords: CBE, FRS, ATM, security challenge, card and currency fraud, MATLAB, Face Recognition.