Objective　To explore the early laboratory indicators for risk of intravenous immunoglobin-resistant Kawasaki diseases. Methods　The clinical data were retrospectively analyzed in 881 Kawasaki disease patients (group A: 26 cases of intravenous immunoglobin-resistant; group B: 855 cases of intravenous immunoglobin-sensitive) from July 1, 2015 to June 30, 2016. After 1:3 matching with age and sex, the regression model for each of variables including sex, age, fever days, temperature, red blood cell count (RBC), white blood cell count (WBC), neutrophil (N), lymphocyte (L), platelet count (PLT) and C reactive protein (CRP), was constructed by conditional logistic regression analysis. Results　Compared with group B, group A had significantly lower RBC count and higher PLT (P<0.05). Logistic regression analysis showed that, with the age, the regression model was Y=－2.87+0.01×PLT (PLT OR=1.01, 95%CI: 1.00~1.01, P<0.01); with the sex, Y=－32.98+0.44×WBC+0.28×          N+0.01×PLT (WBC OR=1.55, 95%CI: 1.17~2.05, P<0.01; N% OR=1.32, 95%CI: 1.04~1.68, P<0.05; PLT OR=1.01, 95%CI 1.00~1.01, P<0.01). Conclusion　In case that abnormally high levels of PLT exist in confirmed Kawasaki disease, it should be aware of possibility of the intravenous immunoglobin-resistant Kawasaki disease.