Atopic dermatitis forms an active area of basic and clinical research, where important new knowledge about genetics and immunopathogenesis has surfaced over the past years, and where simultaneous development of new and innovative therapies is under way. However, the inclusion of any patient in an atopic dermatitis study, whether it is on its genetics, pathogenesis or therapy, requires a diagnosis which is irrefutable. Since there is no simple and also no complicated laboratory procedure to reach a diagnosis of atopic dermatitis, different sets of clinical criteria have been developed for the purpose of making the diagnosis uniformly in different studies as well as in different study centers. The most commonly used are Hanifin and Rajka's set of diagnostic features, which have major and minor clinical criteria to be fulfilled in order to establish a diagnosis of atopic dermatitis. Recent developments in the immunology of atopy have clearly established the major abnormality in this syndrome, the preferential production of allergen-specific IgE. In this contribution, it is suggested that the presence of such antibodies in a given patient should be a mandatory criterium for the diagnosis of atopic dermatitis. Such a diagnostic test however establishes a diagnosis of atopic syndrome, not atopic dermatitis. Thus, for atopic dermatitis we have to rely, for the time being, on additional clinical criteria. The clinical features described in the literature are critically evaluated, and it is suggested that in addition to the mandatory presence of allergen-specific IgE, 2 of 3 principal criteria (pruritus, typical morphology and distribution, chronic or chronically relapsing) should be present for such a diagnosis. Finally, the minor features originally described by Hanifin and Rajka and later evaluated by others are revised and divided over 4 subcategories; a) related to subclinical eczema; b) related to dry skin; c) extra skin folds; and d) ophthalmological pathology. They are suggested to be used as additional criteria only, needed when clinical suspicion is high but the new mandatory and principal diagnostic criteria described here are inconclusive. For study purposes, we suggest that the mandatory and principal criteria are sufficient. They are now evaluated and validated in ongoing atopic dermatitis treatment studies.