a b s t r a c tThe aim of this study was to characterize the adaptive profile and identify variables with great discriminatory power of the Brazilian Azul goat population and Italian Garfagnina population, through the use of principal component and canonical discriminant analysis. A total of 110 Garfagnina milking females (60 in winter and 50 in summer) and 80 Brazilian Azul (40 in winter and 40 in summer) were considered. Air temperature ( • C), black globe temperature (BGT) and relative humidity (%) were measured with the aid of an automatic weather station. Some physiological parameters (rectal temperature -RT, respiratory rate -RR, skin temperature -ST and heart rate -HR), some anatomical parameters (hair diameter -HD and hair length -HL), some hematological parameters (erythrocyte -RBCs, packed cell volume -PCV and mean corpuscular volume -MCV), some blood biochemical parameters (glucose -GLI, cholesterol -COL, triglycerides -TRI, creatinine -CRE, urea -URE, total protein -PRT, albumin -ALB, globulin -GLO, albumin and globulin ratio -A/G, gamma -glutamyl transferase -GGT and aspartate aminotransferase -AST) and some stressed hormones (thyroxine -T4, triiodothyronine -T3 and cortisol -COR) were measured. The variables with greater discriminant power were T3, ST, COR, T4, GGT, HD, GLO, HL and PCV to Garfagnina population and PRT, MCV, PCV, ALB, T4, ST, HL, RBCs, TRI and GGT in the Azul Brazilian population. Classification of the animals was more accurate when considering morphological, physiological, hematological, biochemical and hormonal variables jointly.