Sera from 24 patients with dermatitis herpetiformis and 80 control subjects (patients with other bullous diseases, nonbullous dermatoses, and noncutaneous diseases) were studied to determine the usefulness of assay for IgA antiendomysial antibodies (IgA-EMA) in the diagnosis of dermatitis herpetiformis. The overall sensitivity of IgA-EMA for the diagnosis of dermatitis herpetiformis was 79% and the specificity was 96%. When the three patients with dermatitis herpetiformis who were faithfully following gluten-free diets were excluded, the sensitivity was 90% and the specificity was 96%. No patient in the bullous disease control group (including patients with linear IgA bullous dermatosis) had circulating IgA-EMA. One patient, who did not have direct immunofluorescence evidence for dermatitis herpetiformis but had IgA nephropathy, had a positive IgA-EMA result, an interesting association in light of the rare reports of dermatitis herpetiformis in patients with IgA nephropathy and IgA antigliadin antibodies associated with IgA nephropathy. Although direct immunofluorescence testing of skin biopsy specimens remains the most definitive diagnostic test for dermatitis herpetiformis, indirect immunofluorescence assay of serum for IgA-EMA is a minimally invasive study with a high sensitivity and specificity for dermatitis herpetiformis.