The aim of this paper is to predict different types of pathological subjects from a population through the physical observational variables (HB, RBC, MCH, WBC, MCV, HCT, MCHC, PLT), sex, and age. To achieve this, we have used a multilinear regression model. The obtained results in terms of the proposed model are seen to be accurate enough. Thus, the regression model is expected to provide important preliminary information to the clinicians in order to plan appropriate treatment programs for their patients. This work is carried out in terms of the dataset consisting of 539 subjects provided from blood laboratories in Iraq. The resulting model is the regression model consisting of observations (the blood variables, age, and sex) and the type of anemia. Therefore, the model is developed for predicting the type of anemia to help for diagnosis of these diseases.