BACKGROUND AND OBJECTIVE: The selection of allergen panels is a prerequisite to effectively test for innumerable allergens scattered throughout the environment. However, the selection of the pre-existing panel has been vague and contains some allergens that have not been verified as being common in Korea. This study was aimed to produce allergen panels in Korea. METHODS: For 12 months in 1996, sera were tested by the chemiluminescent assay of Multiple allergen simultaneous test (MAST-CLA: Immunosystems, Mountain view, U.S.A.). A total of 2, 467 specimens that either tested positive or were negative but had high total IgE level were pooled together. The pooled ser a were assayed for 60 allergens supplied by Dexall Acti Tip System (Dexall biomedical Labs. Inc., Gaithersburg, U.S.A.), a recently developed enzyme immunoassay. According to the Allerg Ens Unit (Allergen Unit:AU), 12 of the most frequently encountered and 6 of the leaot frequent allergens with reactions between classes 3 and trace were selected. RESULTS: The 12 most frequently encountered allergens were : Dermatophagoides pteronyssinus, Dermatophagoides farinae, house dust, timothy grass, perennial rye, mugwort, birch, oak, hazel nut, common ragweed, alder and dog dander. The 6 least frequently encountered were : wheat, egg-white, cat epithelium, milk, cockroach and shrimp. CONCLUSION: The 12 allergens we chose proposed to be the minimally required panel of frequently encountered allergens in allergy testing. We conclude that the 12 most frequent allergens should be tested with the total IgE level as a major panel (panel-M) and that the 6 least frequently encountered allergens may be tested separately when needed, as a minor panel (panel-m).