We conducted this systematic review and meta‐analysis to evaluate the existing evidence and to quantitatively synthesise evidence on the impact of therapeutic plasma exchange (TPE) on severe COVID‐19 patients. This systematic review and meta‐analysis protocol was prospectively registered on PROSPERO (CRD42022316331). We systemically searched six electronic databases (PubMed, Scopus, Web of Science, ScienceDirect, http://clinicaltrial.gov, and Cochrane Central Register of Controlled Trials) from inception until 1 June 2022. We included studies comparing patients who received TPE versus those who received the standard treatment. For risk of bias assessment, we used the Cochrane risk of bias assessment tool, the ROBINS1 tool, and the Newcastle Ottawa scale for RCTs, non‐RCTs, and observational studies, respectively. Continuous data were pooled as standardized mean difference (SMD), and dichotomous data were pooled as risk ratio in the random effect model with the corresponding 95% confidence intervals (CI). Thirteen studies (one randomized controlled trials (RCT) and 12 non‐RCTs) were included in the meta‐analysis, with a total of 829 patients. There is a moderate‐quality evidence from one RCT that TPE reduces the lactic dehydrogenase (LDH) levels (SMD −1.09, 95% CI [−1.59 to −0.60]), D‐dimer (SMD −0.86, 95% CI [−1.34 to −0.37]), and ferritin (SMD −0.70, 95% CI [−1.18 to −0.23]), and increases the absolute lymphocyte count (SMD 0.54, 95% CI [0.07–1.01]), There is low‐quality evidence from mixed‐design studies that TPE was associated with lower mortality (relative risk 0.51, 95% CI [0.35–0.74]), lower IL‐6 (SMD −0.91, 95% CI [−1.19 to −0.63]), and lower ferritin (SMD −0.51, 95% CI [−0.80 to −0.22]) compared to the standard control. Among severely affected COVID‐19 patients, TPE might provide benefits such as decreasing the mortality rate, LDH, D‐dimer, IL‐6, and ferritin, in addition to increasing the higher absolute lymphocyte count. Further well‐designed RCTs are needed.