HS-SPME analysis of terpenes does usually have inherent quantification problems when working with complex samples, especially due to the matrix effect of the substrate or the calibration solution. Three different terpene carrier matrices were compared: methanol, synthetic oil and milk fat obtained by centrifugation from milk cream. Considerable differences in calibration sensitivity parameters were observed depending on the matrix used and on the type of terpene standard analysed. For milk sample quantification purposes internal standard method was preferred using milk fat as calibration matrix. Linearity range, repeatability, recovery and limits of detection and quantification were determined. Validation parameters were different depending on the concentration and molecular structure of each terpene analysed, particularly between mono- and sesquiterpenes. The method was useful to determine in an accurate manner the terpene content in milk samples from pasture fed animals, and it will help to establish objective terpene levels to differentiate milks from specific production systems.