Objective. To investigate the validity of the global APS score (GAPSS) to predict thrombosis in patients with autoimmune diseases.Methods. This prospective cohort study included consecutive patients with aPL or SLE. aPL, aPS-PT and GAPSS were determined. A Cox proportional hazards model assessed the validity of GAPSS and identified other potential independent predictors of thrombosis.Results. One hundred and thirty-seven patients [43.5 (S.D. 15.4) years old; 107 women] were followed up for a mean duration of 43.1 (S.D. 20.7) months. Mean GAPSS was significantly higher in patients who experienced a thrombotic event compared with those without [10.88 (S.D. 5.06) vs 8.15 (S.D. 5.31), respectively, P = 0.038]. In univariate analysis, age [hazard ratio (HR) = 1.04 (95% CI 1.01, 1.08)] and GAPSS above 16 [HR = 6.86 (95% CI 1.90, 24.77)] were each significantly associated with thrombosis during follow-up, while history of arterial thrombosis [HR = 2.61 (95% CI 0.87, 7.82)] failed to reach significance. Among aPL assays, IgG aPS/PT—a component of the GAPSS—was significantly associated with thrombosis [HR = 2.95 (95% CI 1.02, 8.51)]. In multivariate analysis, GAPSS above 16 remained the only significant predictor of thrombosis [HR = 6.17 (95% CI 1.70, 22.40)].Conclusion. This first external validation study confirmed that GAPSS can predict thrombosis in patients with aPL and associated autoimmune diseases.