Serum IgA-antigliadin antibodies (SAGA) were measured by ELISA in 46 children with proven celiac disease (CD), in 52 children with probable CD, and in 85 control subjects. Small intestinal biopsy was done within 3 months of the SAGA determination in all children. In the proven and probable CD groups, SAGA values were greater than 70 mu/ml in 76 of 82 biopsies that showed severe mucosal damage, but in only 4 of 31 with normal mucosa; thus specificity was 87.1% and sensitivity 92.7%. In the control group, only 12 of 85 children with normal biopsies had a similarly raised SAGA value. SAGA levels decreased significantly when a gluten-free diet was instituted, and rose to abnormal levels in most children on gluten challenge. IgA-SAGA can be used for monitoring dietary compliance during different phases of CD. Although it may help in selecting some children who are evaluated for the possibility of CD, for small intestinal biopsy, children with active CD may show negative SAGA values. The test should therefore not be used as the sole or final determining factor for the performance of small intestinal biopsy. In the proper clinical setting, a biopsy should be done, regardless of the SAGA results.