Acid-base disorders were studied in two groups of dairy cows. According to conventional interpretation in the 1st group of dairy cows (n = 10) the results of the acid-base parameters indicated respiratory alkalosis and in the 2nd group (n = 10) metabolic acidosis. The data of the two groups were examined using principal component analysis. So called "reduced variables"--principal components for each animal were calculated. Each experimental animal was projected in the coordinates of 3 principal components K1, K2 and K3. The components K1, K2 and K3 comprised a 90% data variability. The importance of variables for a dividing of groups (1 and 2) was quantified. The most important were pCO2, (HCO3)' and BE. It was proved according to mathematico-empiric equations that dairy cows of 1st group suffered from acute respiratory alkalosis. In the 2nd group of dairy cows the combined acid-base disorder was confirmed using the equations for metabolic acidosis and chronic respiratory acidosis. The use of PCA method and empiric equations for acid-base imbalances shows the possibilities for gaining new knowledge how to make the diagnostic process more exact.