The substitution of bovine milk for ovine and caprine milk in cheesemaking is a real economic and analytical problem in dairy industry. A capillary electrophoresis method was employed for the simultaneous quantitative determination of bovine, ovine and caprine casein fractions from a Mexican unripened cheese (Panela). Partial least squares regressions (PLS-1 and PLS-2), principal component regression (PCR) and multiple linear regression (MLR) were compared to build calibration models for the prediction of bovine, ovine and caprine milk percentages employed to manufacture Panela cheeses. The chemometric methods were prepared by measuring the peak areas of bovine β-casein A 1, β-casein A 2, κ-casein, α s1-9P-casein and α s1-8P-casein; ovine β 1-casein, β 2-casein, κ-casein and α s1-8P-caseins (I and II); and caprine β 1-casein, β 2-casein and κ-casein from the training set of cheeses. The given models were validated by applying them to the analysis of a prediction set of cheeses. It was found that MLR led to more precise predictions than the other multivariate calibration methods with a root square error under 2.2%.