To investigate the relationship between short-term changes in quantitative myasthenia gravis score (QMGS) after thymectomy and postoperative recurrence in myasthenia gravis (MG) patients without thymoma.A retrospective observational cohort study. The QMGS of 44 patients with non-thymomatous MG were evaluated before and 1 month after thymectomy, and the frequency and time of postoperative recurrence were recorded. The reduction rate of QMGS (rr-QMGS) was defined as (QMGS one week before thymectomy - QMGS one month after thymectomy)/ QMGS one week before thymectomy × 100 %, as an indicator of short-term symptom change after thymectomy. The receiver operating characteristic (ROC) curve was established to determine an appropriate cut-off value of rr-QMGS for distinguishing postoperative recurrence. Multivariate Cox regression analysis was applied to predict postoperative recurrence.Postoperative recurrence occurred in 21 patients (30 times in total) during follow-up. The mean annual recurrence rate was 3.98 times/year preoperatively and 0.30 times/year postoperatively. ROC analysis determined the cut-off value of rr-QMGS was 36.7 % (sensitivity 90.5 %, specificity 52.2 %). Multivariate Cox regression analysis showed that rr-QMGS＜36.7 % (hazard rate[HR]6.251, P = 0.014) is positive predictor of postoperative recurrence. Kaplan-Meier analysis showed that postoperative recurrence time was earlier in the low rr-QMGS group than in the high rr-QMGS group (12.62 vs. 36.60 months, p = 0.005).Low rr-QMGS is associated with early postoperative recurrence. Rr-QMGS can be used to predict postoperative recurrence of non-thymomatous MG.Copyright © 2024. Published by Elsevier Ltd.