Whether paroxysmal nocturnal hemoglobinuria (PNH) clone in aplastic anemia (AA) is a prognostic factor to immunosuppressive therapy is a subject of debate. We evaluated hematological responses to immunosuppressive therapy (IST) in severe AA (SAA) patients with or without the presence of a PNH clone. In 97 SAA patients who received first-line IST between January and December 2011, 24 (24.7 %) had a PNH clone prior to treatment, with a median clone size of 7.82 % (range 1.19-45.46 %). The response rates to IST for patients with or without a PNH clone were 66.7 and 50.7 % (P < 0.172), 79.2 and 57.5 % (P < 0.057), and 79.2 and 67.1 % (P < 0.264) at 3, 6, and 12 months, respectively. Combined rate of complete and good partial responses differed between patients with or without a PNH clone: insignificantly at 3 months (41.7 vs. 21.9 %, P < 0.058), but significantly at 6 (66.7 vs. 31.5 %, P < 0.002) and 12 (75.0 vs. 46.6 %, P < 0.015) months. Multivariate analysis revealed that a pretreatment neutrophil count of >0.2 × 10(9)/L is indicative of a better response, while the presence of a PNH clone is predictive to a higher combined rate of complete and good partial responses. This study demonstrated that the presence of a PNH clone could predict a better hematological response instead of a higher response rate in patients with SAA.