Previous studies reported the positive viral RNA among coronavirus disease-2019 (COVID-19) recovered patients. This study aimed to summarize the current evidence of factors associated with the risk of disease recurrence. PubMed and Embase were searched until September 2020 to identify studies assessing characteristics of recurrence and nonrecurrence subjects after discharge. Random-effect meta-analysis was used to pool estimates of odds ratio (OR) or weighted mean difference (WMD) and their 95% confidence intervals (CIs) across studies. Meta-analysis data were available for age, sex, hospital duration, disease severity, seven comorbidities, five symptoms, five indexes of blood routine, nine indexes of blood biochemistry, four treatment therapies, two antibodies, and history of high-risk contact. Among them, hospital duration of recurrence cases was significantly shorter than nonrecurrence subjects (WMD, -1.55 days; 95% CI, -2.66 to -0.45). Fatigue, positive Immunoglobulin M (IgM), and positive IgG were associated with an increased risk of recurrence cases, with ORs and 95% CIs of 4.06 (1.14-14.4), 2.95 (1.15-7.61), and 3.45 (1.58-7.54), respectively. In contrast, the odds of recurrence cases were observed to significantly lower in subjects with elevated lactate dehydrogenase and C-reactive protein, low lymphocyte count, steroid and arbidol use, with ORs (95% CIs) of 1.08 (0.27-4.37), 0.49 (0.27-0.97), 0.64 (0.42-0.97), 0.48 (0.25-0.96), and 0.48 (0.25-0.92), respectively. This study provided up-to-date evidence of several clinical and epidemiological characteristics in the association with COVID-19 recurrence cases. Further in-depth analyses for the causal effect of factors on re-positive viral RNA are needed for the management of discharged patients with COVID-19.