80 points) according to the PCIS. Serum levels of copeptin, TLR2, and TLR4 were compared in each group. The relationship between PCIS score and serum levels of copeptin, TLR2 and TLR4; serum levels of copeptin, TLR2, and TLR4 in children with different prognosis in the severe group; and efficacy of serum copeptin, TLR2 and TLR4 levels in predicting prognosis of severe patients were compared. Results The serum levels of copeptin, TLR2 and TLR4 in group A, B, C and the mild disease group were all higher than those in the healthy group (P<0.05). Serum levels of copeptin, TLR2 and TLR4 in group A, B and C were all higher than those in the mild group (P<0.05). The serum levels of Copeptin, TLR2 and TLR4 were significantly different among groups A, B and C (P<0.05). The PCIS scores of children in the severe disease group were negatively correlated with serum copeptin, TLR2 and TLR4 levels, and the correlation coefficients (r) were -0.831, -0.885, and -0.907, respectively (P<0.05). The mortality rate in the severe group was 6.09% (22/361), and the serum levels of copeptin, TLR2 and TLR4 in the surviving children were all lower than those in the dead children (P<0.05). The best cut-off points of serum copeptin, TLR2, and TLR4 levels for predicting severe pneumonia in children were 1.85 ng/mL, 11.55 ng/mL, and 7.18 ng/mL, respectively. The sensitivity, specificity and AUC of the combination were 90.91%, 97.87%, and 0.869, respectively. Conclusion The serum levels of copeptin, TLR2 and TLR4 in children with severe pneumonia caused by mycoplasma pneumoniae infection are generally high. It is suggested that the dynamic changes of serum levels of copeptin, TLR2 and TLR4 should be regularly detected in the treatment of children with severe pneumonia caused by mycoplasma pneumoniae infection, so as to assess the condition and physiological status, guide the treatment and predict the recent prognosis."/>