Milk and milk products are one of the most consumed foods. However, there has been an increase in milk adulteration, especially goat milk. Current methods for detection of milk adulteration are expensive, impractical and are not sufficient to answer quantitatively. Therefore, to meet this demand, a rapid, easy to use and inexpensive method was developed in this study. The proposed methodology is based on Fourier Transform Infrared Spectroscopy (FTIR) with the combination of multivariate calibration techniques for determination of adulteration of goat milk with cow milk. During the study, 150 raw goat milk samples were collected from different goats in 3 different seasons (June-2014, December 2014, March2015). Then, adulterated samples were prepared with goat milk, cow milk and water. Both adulterated and raw goat milk samples were analyzed with FTIR. Afterwards, 7 different multivariate calibration models (for each season (3), binary combination of seasons(3) and a ternary combination of all the seasons) were generated with synthetically prepared samples by using Genetic Inverse Least Squares (GILS) and Partial Least Squares (PLS) methods. These models were used to predict both adulterated and raw goat milk samples in order to evaluate the success of the models. Standard Error of Prediction (SEP) values for goat milk, cow milk and water are 7.9, 6.3 and 4.9 (w/w %), respectively, indicate satisfactory predictions by GILS. On the other hand, PLS models gave SEP values ranging between 6.4 and 12.9 (w/w %).