Serological testing to diagnose Helicobacter pylori infection in children is still controversial, although commonly used in clinical practice. We compared the immunoglobulin G (IgG) and IgA results of two commercially available enzyme immunoassays (EIAs) (Pyloriset IgG and IgA and Enzygnost II IgG and IgA) for 175 children with abdominal symptoms divided into three age groups (0 to 6 years, n = 47; >6 to 12 years, n = 77; >12 years, n = 51). A child was considered H. pylori infected if at least two of three tests (histology, rapid urease test, ¹³C-urea breath test) or culture were positive and noninfected if all results were concordantly negative. Of 175 children, 93 (53%) were H. pylori negative and 82 were H. pylori positive. With the recommended cutoff values, the overall specificity was excellent for all four EIAs (95.7 to 97.8%) regardless of age. Sensitivity varied markedly between tests and was 92.7, 70.7, 47.5, and 24.4% for Enzygnost II IgG, Pyloriset IgG, Enzygnost II IgA, and Pyloriset IgA, respectively. Sensitivity was low in the youngest age group (25 to 33.3%), except for Enzygnost II IgG (91.6%). Receiver-operating curve analyses revealed that lower cutoff values would improve the accuracy of all of the tests except Enzygnost II IgG. Measurement of specific IgA, in addition to IgG, antibodies hardly improved the sensitivity. The specificity of commercial serological tests is high in children when the cutoff values obtained from adults are used. In contrast, sensitivity is variable, with a strong age dependence in some, but not all, tests. We speculate that young children may have a different immune response to H. pylori, with preferable responses to certain antigens, as well as lower titers than adults. The Pyloriset test may fail to recognize these specific antibodies.