Hypereosinophilia is a common biological finding, arising in a number of different clinical situations. Hypereosinophilia can result from the presence of a defect in the hematopoietic stem cell giving rise to a primary eosinophilia, it can be present in many myeloproliferative disorders or alternatively it may be a reactive form, secondary to many clinical conditions. The hybrid gene FIP1L1-PDGRFa was identified in a subset of patients presenting with primary HES. In spite of this, the majority of HES patients do not present detectable molecular lesions. So far, for the majority of them, the diagnosis is based on exclusion criteria and in same cases it may remain doubt. Since many therapeutic approaches now available may change the prognosis of primary HES patients, it is actually of great clinical importance to identify these patients. The aim of the study was to explore the possibility to distinguish between primary and secondary hypereosinophia basing on WT1 expression. BM and PB samples from 292 patients affected by hypereosynophilia and 50 BM and PB from healthy subjects were studied. In 202 patients (69%) clinical and/or laboratory findings were consistent with other eosinophilic disorders, including Churg-Strauss vasculitis, chronic eosinophilic pneumonia, eosinophilic gastroenteritis and episodic angioedema. 90 patients (31%) were diagnosed as primary HES or chronic eosinophilic leukaemia (CEL). Conventional cytogenetic analysis and FISH were performed to study the involvement of the PDGFRalpha and beta and FGFR1. All the patients were characterized at the molecular level for the presence of the known fusion transcript associated with eosinophilia: BCR-ABL, CBF-MHY11, AML-ETO, BCR-FGFR, TEL-PDGFR and for FIP1L1-PDGFRa transcript. In primary HES patients treated with Imatinib (70 out of 90) or with other conventional therapies (hydroxiurea in 8 pts, IFNalpha in 4 pts) the BM and PB samples were analyzed at regular time intervals during treatment (after 1, 3, 4, 6,12 months). 32 patients were positive for the presence FIP1L1-PDGFRa. Real Time PCR for the detection of WT1 transcript amount was performed at diagnosis and during follow-up in all the 292 patients. We found that WT1 amount in secondary HES is not significantly different compared to healthy controls with a mean value of WT1 copies/104 ABL copies of 32 (range 0–75) in BM and 5 (range 0–17) in PB vs 35 (range 3–90) in normal BM and 4 (range 1–20) in normal PB. By contrast, primary HES showed significantly higher values compared to normal controls or to secondary conditions with a mean of 651 WT1 copies in BM (p=0,0003) and 151 in PB (p=0,00002). WT1 is not significantly different in FIP1L1-PDGFRa when compared to HES with normal karyotype (p=0,3). Moreover, WT1 transcript follows the behaviour of FIP1L1-PDGFRa or of the specific cytogenetic marker during follow-up. In conclusion the quantitative assessment of WT1 is a useful tool for the diagnosis and follow-up of primary HES patients with normal karyotype.