This study explored circulating pneumoproteins in the diagnosis, severity, and prognosis of COVID-19 by meta-analysis. We searched six English databases until December 16, 2021. Standardised mean difference (SMD) and 95% confidence interval (CI) were the overall outcomes. RevMan 5.3, Stata 16, and Meta-DiSc 1.4 were utilised for pooled analyses. Totally, 2902 subjects from 29 studies were included. COVID-19 patients had higher circulating KL-6, SP-D, and SP-A levels (SMD=1.34, 95% CI [0.60, 2.08]; SMD=1.74, 95% CI [0.64, 2.84]; SMD=3.42, 95% CI [1.31, 5.53], respectively) than healthy individuals. Circulating KL-6 levels were lower in survivors than in non-survivors (SMD=-1.09, 95% CI [-1.63, -0.55]). Circulating KL-6, SP-D, and RAGEs levels in mild to moderate patients were significantly lower (SMD=-0.93, 95% CI [-1.22, -0.65]; SMD=-1.67, 95% CI [-2.82, -0.52]; SMD=-1.17, 95% CI [-2.06, -0.28], respectively) than severe patients. Subgroup analysis and meta-regression suggested that age may be responsible for the heterogeneity (P=0.071) when analysing KL-6 in mild to moderate vs. severe patients. The meta-analysis of diagnostic accuracy including KL-6 for severity and mortality, and SP-D for severity demonstrated that they all had limited diagnostic values. Therefore, circulating pneumoproteins (KL-6, SP-D, and RAGEs) may reflect the diagnosis, severity, and prognosis of COVID-19 to some extent.