The updated Sapporo classification criteria for antiphospholipid syndrome (APS) only include thrombosis or pregnancy morbidity as clinical criteria. To test this notion, we studied 55 patients (80% women) with hematologic manifestations. All fulfilled the laboratory criteria for primary APS. Thirty-five patients (64%) had thrombocytopenia, 14 (25%) had autoimmune hemolytic anemia, and 6 (11%) had both. Twenty-five patients (22 women, 88%) also fulfilled one clinical criterion for APS after a median followup of 13.2 years (range, 1.45-37 years), whereas the remaining 30 patients (22 women, 73%) have not had any thrombotic event nor pregnancy morbidity after a median follow-up of 5.4 years (range, 0.12-24 years). No patient developed systemic lupus erythematosus during followup. The hematologic manifestation was asynchronous with the APS onset in 84% of patients. The response to treatment was similar regardless of the APS status. Patients with definite APS were more frequently positive for the lupus anticoagulant (63%) than lupus anticoagulantpositive patients without APS (30%; odds ratio, 3.5; 95% confidence interval, 1.07-11.4; P < .02). Anticardiolipin or anti- 2 -glycoprotein-I antibodies were highly prevalent among the study groups. Our study suggests that, depending upon their antiphospholipid profile, patients with hemocytopenias appear to comprise a peculiar subset of patients with APS; some develop thrombotic and/or obstetric APS whereas others continue with hematologic APS. (Blood. 2010;116(16):3058-3063)