OBJECTIVE The aim of this study was to evaluate the clinical relevance of the global APS score (GAPSS) in a cohort of primary APS patients. METHODS This study included 62 consecutive patients with primary APS. Data on clinical manifestations, conventional cardiovascular risk factors and aPL profile were collected. The GAPSS was calculated for each patient by adding together the points corresponding to the risk factors, based on a linear transformation derived from the β regression coefficient as follows: 3 for hyperlipidaemia, 1 for arterial hypertension, 5 for aCL IgG/IgM, 4 for anti-β2 glycoprotein I IgG/IgM, 3 for aPS-PT IgG/IgM and 4 for LA. RESULTS Higher GAPSS values were seen in patients who experienced thrombosis alone when compared with those with pregnancy loss alone [11.5 (S.D. 4.6) and 8.7 (S.D. 3.2), P = 0.04]. Patients with both thrombosis and pregnancy loss showed higher GAPSS than those with pregnancy loss alone [12.5 (S.D. 4.6) vs 8.7 (S.D. 3.2), P = 0.02]. Higher GAPSS values were also shown after subgrouping for the site of thrombosis when compared with pregnancy loss alone [12.2 (S.D. 5.2) for arterial thrombosis, 12.0 (S.D. 4.0) for venous vs 8.7 (S.D. 3.2), P = 0.02 and P = 0.04, respectively]. Patients with thrombotic recurrences showed higher GAPSS values when compared with those without recurrence [13.7 (S.D. 3.1) vs 9.4 (S.D. 3.9), P = 0.02]. This was also seen when comparing recurrences vs no recurrences independently of the site of the thrombotic event [13.9 (S.D. 3.6) vs 11.0 (S.D. 4.3), P = 0.01 for arterial and 13.6 (S.D. 2.18) vs 8.91 (S.D. 3.6), P < 0.01 for venous thrombosis]. GAPSS values ≥11 were strongly associated with a higher risk of recurrence [odds ratio (OR) 18.27 (95% CI 3.74, 114.5) for a cut-off of 11, OR 20.64 (95% CI 3.92, 185.92) for a cut-off of 12 and 21.64 (95% CI 3.89, 189.56) for a cut-off of 15]. GAPSS values ≥11 seemed to have the best risk accuracy in terms of sensitivity and specificity. CONCLUSION The GAPSS is demonstrated to be a valid tool for a substantial improvement in risk stratification for thrombosis in primary APS.