Sensory assessment of virgin olive oil (“panel test”) is the only sensory method included in international regulations of edible oils and its application is compulsory. Even if its application has been a success in quality control, improving the quality of virgin olive oils over the last 30 years, at present, there is no reference material (RM), in the strict sense of the term, to be used as a validated standard for sensory defects of virgin olive oil with which tasters can be trained. Usually, real samples of virgin olive oils assessed by many panels for the International Olive Council (IOC) ring tests are used as materials of reference in panel training and control. The latter are highly representative of the main perceived defects, but availability is limited, samples are not homogeneous year after year, and other secondary defects can be present. Thus, in order to provide solutions, this work describes an analytical procedure for implementing olfactory formulations that emulate rancid and winey-vinegary defects found in virgin olive oils with the aim of providing reproducible RMs that can be prepared on demand. A strategy for designing RMs for aroma is presented and the optimization process to obtain the best formulation is described. Under the criteria of representativeness, verified with the advice of the IOC, aroma persistence, and simplicity in formulation, two RMs for winey-vinegary and rancid were obtained by diluting acetic acid and ethanol (winey-vinegary defect) and hexanal (rancid defect) together with other compounds that are used to modify aroma and avoid non-natural sensory notes.