The clinical accuracy of the Immulite-2000 for latex-specific IgE was compared to 3 other FDA-cleared immunoanalyzers. We compared the clinical accuracy by ROC of the DPC microplate AlaSTAT (Ala), Hycor HY-TEC (HT), Pharmacia CAP System and Immulite-2000). Stored sera (201) from subjects in a multicenter latex skin testing study (J Allergy Clin Immunol 1998;102:482-90) classified as having positive [n = 84] or negative [n = 117] latex puncture skin tests [PST, non-ammoniated latex]) were evaluated by PST and latex specific IgE levels. Using the manufacturers' positive cutoff of > 0.35 kU/L latex-specific IgE, ROC areas under the curve (AUCs) + SE were: CAP 0.870 + 0.030; Ala 0.847 + 0.032; HT (> 0.05 kU/L) 0.913 + 0.024; and Immulite-2000, 0.880 + 0.029. No significant differences were detected between AUCs (P>0.05) for the assays at > 0.35 kU/L. When the Immulite-2000 was used at > 0.10 kU/L (Imm 0.1), the AUC was 0.919 + 0.024, significantly greater than Ala and Immulite-2000 (P<0.05). Cutoff concentrations for (maximal diagnostic efficiencies, MDE) were > 0.35 kU/L for Imm (89.6%), CAP (88.1%) and Ala (86.6%). HT and Imm 0.1 had MDEs at > 0.15 kU/L and > 0.10 kU/L of 87.6% and 92.1%, respectively. The MDE of Imm 0.1 was greater (P<0.05) than Ala, HT and CAP, but equivalent to Immulite-2000 (P>0.05). In this sample cohort, the Immulite-2000 and Imm 0.1 are the most accurate in vitro assays for detecting latex-specific IgE in serum when using latex PST as a discriminator.