Kawasaki disease (KD) is a form of systemic vasculitis that generally occurs in children under 5 years old. Currently, KD is still diagnosed according to its clinical symptoms, including prolonged fever, skin rash, conjunctivitis, neck lymphadenopathy, palm erythema, and oral mucosa changes. Because KD is a type of inflammation without specific marker for diagnosis, we plan to profile the plasma antibodies by using E. coli proteome microarray and analyze the differences between KD and healthy subjects. Plasmas were collected from KD patient before intravenous immunoglobulin treatment (KD1), at least 3 weeks after treatment (KD3), nonfever control (NC), and fever control (FC) children. The initial screening, which consisted of 20 KD1, 20 KD3, 20 NC, and 20 FC, were explored using E. coli proteome microarrays (~4200 unique proteins). About ~70 proteins were shown to have high accuracy, e.g. 0.78~0.92, with regard to separating KD1, KD3, NC, and FC. Those proteins were then purified to fabricate KD focus arrays for training ( n = 20 each) and blind-testing ( n = 20 each). It only took 125 pl of plasma, less than a drop of blood, in the focus array assays. The AUC scores for blind tests of KD1 versus NC (17 protein markers), KD1 versus FC (20 protein markers), KD3 versus NC (9 protein markers), and KD1 versus KD3 (6 protein markers) were 0.84, 0.75, 0.99 and 0.98, respectively. This study is the first to profile plasma antibodies in KD and demonstrate that an E. coli proteome microarray can screen differences among patients with KD, nonfever controls, and fever controls.