Introduction: The laboratory diagnosis of paroxysmal nocturnal hemoglobinuria (PNH), disease that is categorized by reduced synthesis of glycosylphosphatidylinositol (GPI) anchor, is based on the detection of blood cells deficient in GPI-anchored proteins by flow cytometry. PNH clones have been detected in patients with aplastic anaemia (AA) and myelodysplastic syndrome (MDS), with therapeutic implications. Objectives: To validate a sensitive assay for detection of GPI-anchored protein-deficient cells, by flow cytometry, and to analyze the clone frequency in AA and MDS patients. Methods: Samples from 20 AA patients, 30 MDS patients and 20 adult volunteers (control group) were analyzed using monoclonal antibodies to CD16, CD24, CD55 and CD59 (neutrophils); CD14 and CD55 (monocytes); CD55 and CD59 (erythrocytes); besides fluorescent aerolysin reagent (FLAER) (neutrophils and monocytes) and lineage markers. The proportions of PNH cells detected in neutrophils and monocytes, using different reagent combinations, were compared by analysis of variance (ANOVA) and Pearson's correlation. Results: PNH cells were detected in five (25%) AA patients, and the proportions of PNH cells varied from 0.14% to 94.84% of the analyzed events. PNH cells were not detected in the MDS patients. However, by the analysis of these samples, it was possible to identify the technical challenges caused by the presence of immature and dysplastic circulating cells. FLAER showed clear distinction of GPI-deficient cells. Conclusion: Multiparameter flow cytometry analysis offers high sensitivity and accuracy in the detection of subclinical PNH clones. FLAER shows excellent performance in detection of PNH neutrophils and monocytes.