Cervix cancer screening nowadays, which is done manually by the cytologist, is inefficient and time consuming. It requires high skills and experiences of the cytologists. This requirement leads to a condition where the diagnosis is inherently prone to the human error. Coping with this problem, this research is to introduce an automated diagnosis algorithm for early detection of cervix cancer. The diagnosis algorithm is developed to recognize pattern on 2-dimensional digital cervical cytological image produced from Pap smear slides. Pattern recognition is applied to variables of cell morphology and color intensity. Afterward, measurements and identification of cells into normal and abnormal class is done on the basis of parameters color intensity, N/C ratio, and 2D wavelet approximation coefficients. The automated diagnosis algorithm is intended to improve reliability as well as to reduce time consumption in the diagnosis of cervix cancer. Therefore, this will produce more accurate, faster and less expensive analysis of Pap smear test, which can be used to provide better health service for public community.