We systematically investigated the predictive value of gross extranodal extension (gENE) for differentiated thyroid carcinoma persistence/recurrence.Retrospective study.A tertiary care hospital.This study was divided into 2 groups according to gENE status: the gENE group and non-gENE group. We compared the disease persistence/recurrence rates of these 2 groups in the entire cohort and by individual risk group (intermediate/high risk), analyzed whether gENE was an independent risk factor for disease persistence/recurrence, and explored the impact of gENE-specific features on disease persistence/recurrence.There were 989 patients who satisfied the inclusion criteria: 57 patients in the gENE group and 932 in the non-gENE group. The disease persistence/recurrence rate of the gENE group was higher than that of the non-gENE group in the entire cohort and by individual risk group (P < .05 for each). Unexpectedly, the outcomes of the gENE group with intermediate risk were similar to those of the non-gENE group with high risk (P = .72). For the entire cohort, gENE was an independent predictor for disease persistence/recurrence (odds ratio, 2.89; 95% CI, 1.39-6.00; P = .005). Specific features of gENE (P > .05 for each) were not related to disease persistence/recurrence.Patients with gENE and intermediate risk might be regraded as high risk. Specific features of gENE have no impact on disease persistence/recurrence.