Analytic variability affects the accuracy of measurements of specific IgE antibodies, but the frequency of false results attributable to analytic variability is not well documented. We have monitored the accuracy of the results generated in our laboratory by testing aliquots of positive serum pools and a negative serum pool submitted blindly for the measurement of IgE antibodies to 16 different allergens, including foods; weed, grass, and tree pollens; mites, molds, and epithelia; and Hymenoptera venoms. Positive serum pools were prepared to contain modest amounts of IgE antibodies. Tests were performed by immunometric assays with microcrystalline cellulose allergen immunosorbents. Frank false-positive and false-negative results were very uncommon when binding levels were classified by a ratio-based reporting scheme. False-borderline results were more common. A borderline result is truly equivocal and may or may not indicate the presence of low levels of IgE antibodies. Analytic variability adds uncertainty to the measurement of small quantities of IgE antibodies regardless of the classification scheme used to report test results.