The importance of obtaining and proving that the odor of an extract is representative of the odor of the original product from which it was obtained, before analysis by GC-FID, GC-MS, or GC sniff, is discussed. The sensory methods, such as triangle tests, matching tests, and quantitative analysis, used to determine the representativeness of the odor of an extract are described. Beer extracts obtained by three methods were used to illustrate the interest of the sensory tests. A method using a mixture of XAD resins was proved by sensory analysis to give some extracts with sensory characteristics representative of the particular beers from which they were obtained. Such sensory evaluation of the quality of the aroma of extracts has to be systematically made when sniffing analysis is applied to new types of beer or to different types of products.