We report three observations of food allergy to Penicillium. A systematic approach has allowed the retention of the surprising diagnosis in the first observation that we met. For those following we were content with a diagnosis of strong suspicion, without making a provocation test. The indications of provocation tests are discussed. As in these observations, it is possible that certain food allergens which are considered as rare are in fact more frequent than those reported in the literature; when an allergy does not bother the patient, when exclusion of the food is easy, food allergy is "forgotten" and not reported to the physician.