Wilms' tumor (WT) is the most frequent malignant renal tumor in children. The survival rate is lower in patients with recurrence, and the factors that influence relapse in WT are not fully understood. Loss of heterozygosity on chromosome 16q (LOH 16q) has been reported to be associated with the relapse in WT, but this remains controversial. We performed a meta-analysis to clarify this. PUBMED, EMBASE, and the Cochrane Library were searched up to March 17, 2017. Ten studies involving 3385 patients were ultimately included in the meta-analysis. The meta-analysis showed that LOH 16q was significantly associated with the relapse in WT (relative risk [RR] = 1.74, 95% confidence interval [CI] = 1.43-2.13, P < 0.00001; hazard ratio [HR] = 1.76, 95% CI = 1.38-2.24, P < 0.00001). No significant heterogeneity among studies or publication bias was found. Sensitivity analysis showed omitting one study in each turn could not change the results. Subgroup analysis based on two studies indicated LOH 16q was more effective on elevated replase risk in patients with favorable-histology WT (RR = 2.52, 95% CI = 1.68-3.78, P < 0.00001; HR = 2.99, 95% CI = 1.84-4.88, P < 0.0001) but further work are needed to confirm this. These findings confirm that LOH 16q increased the relapse risk in WT, but more studies are required to further assess the association between LOH 16q and WT relapse among different subgroups.