We challenge 3 prevailing concepts in understanding atopic dermatitis using data from epidemiologic studies. First, we show that although atopy is associated with atopic dermatitis to some degree, its importance is not likely to be a simple cause-and-effect relationship, especially at a population level. Our epidemiologic data do not exclude a contributory role for IgE-mediated immunologic processes, especially in those with existing and severe disease. Second, evidence is presented that does not support a straightforward inverse relationship between infections and atopic dermatitis risk. A link, if present, is likely to be more complex, depending critically on the timing and type of infectious exposure. Third, recent evidence suggests that the risk of subsequent childhood asthma is not increased in children with early atopic dermatitis who are not also early wheezers, suggesting a co-manifestation of phenotypes rather than a progressive atopic march. Collectively, these observations underline the importance of epidemiologic studies conducted at a population level to gain a more balanced understanding of the enigma of atopic dermatitis.