Objective:To investigate the predictive role of dynamic changes of plasma biomarkers in patients with viral and mycoplasma community-acquired pneumonia (CAP).Methods:From January 2020 to June 2020, 141 patients with viral and mycoplasma CAP in People's Hospital of Ningxia Hui Autonomous Region were enrolled. Pneumonia severity index (PSI) scores [grade Ⅰ-Ⅱ(PSI score ≤ 70), grade Ⅲ (PSI score 71-90) and grade Ⅳ-Ⅴ(PSI score ≥ 91)], serum amyloid A (SAA), hypersensitive C-reactive protein (hs-CRP), procalcitonin (PCT), erythrocyte sedimentation rate (ESR) and white blood cell (WBC) on the 1 day after admission were compared between the different pathogens (viral and mycoplasma) or different disease severity. The change in level of SAA, hs-CRP on the third day (Δ 3 d = 1 d-3 d) were compared among different disease outcome groups (patients were divided into improved group, stable group and exacerbation group based on PSI scores or lung CT images on the third day). The change in the level of SAA, hs-CRP on the seventh day (Δ 7 d = 1 d-7 d) were compared among different disease prognosis groups (patients were divided into survival group and death group based on 28-day survival data). The receiver operating characteristic curve (ROC) were drawn to evaluate the value of SAA in the evaluation of disease and prediction prognosis. Results:The level of SAA in mycoplasma group (43 cases) was significantly higher than that in virus group (98 cases) on the 1 day after admission. There were no significant differences in other plasma biomarkers between the two groups. The more severe the illness, the higher the SAA level on the 1 day after admission. The trends of other plasma biomarkers in the two groups were consistent with SAA. The levels of SAA in the patients with exacerbation of the virus group and mycoplasma group (12 cases, 9 cases) were significantly higher than those of the improved group (57 cases, 26 cases) and the stable group (29 cases, 8 cases). SAA increased gradually in the exacerbation group, decreased gradually in the improved group, and slightly increased in the stable group. ΔSAA 3 d were differences among three groups. The change trend of hs-CPR was consistent with SAA. The level of SAA in the death group was higher than that in the survival group on the seventh day. SAA increased in the death group and decreased in survival group with time from hospital admission. There were differences according to ΔSAA 7 d between death group and survival group. The change trend of hs-CPR was consistent with SAA. ROC curve showed that the value of SAA was better than hs-CRP in assessing the severity of patients on admission day, and the area under ROC curve (AUC) was respectively 0.777 [95% confidence interval (95% CI) was 0.669-0.886], 0.729 (95% CI was 0.628-0.830). The value of ΔSAA 3 d was better than SAA on the third day predicting disease trends, and AUC was respectively 0.979 (95% CI was 0.921-1.000), 0.850 (95% CI was 0.660-1.000). hs-CRP on the third day and Δhs-CRP 3 d had no predictive value. Both SAA on the seventh day and ΔSAA 7 d have predictive value for prognosis. AUC was respectively 0.954 (95% CI was 0.898-0.993) and 0.890 (95% CI was 0.689-1.000). SAA on the seventh day and ΔSAA 7 d were better than hs-CRP on the seventh day. Δhs-CRP 7 d have no predictive value. Conclusions:SAA is a sensitive and valuable indicator for CAP patients with viruses and mycoplasma. Dynamic monitoring of SAA can evaluate the patient's progression, prognosis, and assist diagnosis and treatment.