AGVHD may occur in a high proportion of allo-transplanted patients and is one the main causes of death. ECP treatment may successfully be employed in patients refractory or not eligible to high doses steroid therapy. The ECP intensification, as well as the timely start of the courses, are probably fundamental to enhance the probability and the rate of clinical response. Ten aGVHD III-IV patients, 7 males and 3 females, were enrolled. Four were adults and six were children with liver, gut or skin localization. The mean time from aGVHD diagnosis to ECP starting was seven days. Three ECP were performed weekly for three consecutive weeks then two procedures every two weeks were carried out in six months. A mean of 16 (6–29) procedures per patient was carried out, being 4 (1–6) months the mean duration per treatment. Phenotype studies disclosed Treg cells and circulating oligoclonal TCRVB subsets in in two and five patients, respectively. Cyclosporine A was used as primary immunosuppressive drug in six patients or associated to MMF or tacrolimus in two and one patients. One patient received MMF alone and four, Infliximab (3–6 doses). Two patients died for progressive GVHD and in one ECP was stopped for severe CMV reactivation. Seven patients (78%) had fast and significant clinical recovery after the first six, intensified ECP and the mean follow-up observed is of 29 (6–50) months. One child responsive after seven ECP courses died for acute liver failure, one month after the ECP had been concluded One child relapsed and died for NB in the first year post transplantation, two adults are alive with relapsed HL, the others are alive without clinical or molecular signs of relapse. As regards phenotype studies, in two responsive patients, not treated with Cyclosporine A, Treg absolute counts tripled during ECP treatment while oligoclonal TCRVB families disappeared in all responsive patients. Despite the limited number of the subjects treated, the rapid clinical recovery observed in these highly compromised patients enforces the indication to employ intensive ECP courses to accelerate the clinical remission and to reduce the immunosuppression. As regards the possible use of the Treg or the TCRVB oligoclonal subsets modifications as prognostic factors during ECP treatment, the true involvement of these subpopulations is up to date, unclear. In particular Cyclosporine A might inhibit the production of Tregs and probably, other biological markers must be individuated. AGVHD may occur in a high proportion of allo-transplanted patients and is one the main causes of death. ECP treatment may successfully be employed in patients refractory or not eligible to high doses steroid therapy. The ECP intensification, as well as the timely start of the courses, are probably fundamental to enhance the probability and the rate of clinical response. Ten aGVHD III-IV patients, 7 males and 3 females, were enrolled. Four were adults and six were children with liver, gut or skin localization. The mean time from aGVHD diagnosis to ECP starting was seven days. Three ECP were performed weekly for three consecutive weeks then two procedures every two weeks were carried out in six months. A mean of 16 (6–29) procedures per patient was carried out, being 4 (1–6) months the mean duration per treatment. Phenotype studies disclosed Treg cells and circulating oligoclonal TCRVB subsets in in two and five patients, respectively. Cyclosporine A was used as primary immunosuppressive drug in six patients or associated to MMF or tacrolimus in two and one patients. One patient received MMF alone and four, Infliximab (3–6 doses). Two patients died for progressive GVHD and in one ECP was stopped for severe CMV reactivation. Seven patients (78%) had fast and significant clinical recovery after the first six, intensified ECP and the mean follow-up observed is of 29 (6–50) months. One child responsive after seven ECP courses died for acute liver failure, one month after the ECP had been concluded One child relapsed and died for NB in the first year post transplantation, two adults are alive with relapsed HL, the others are alive without clinical or molecular signs of relapse. As regards phenotype studies, in two responsive patients, not treated with Cyclosporine A, Treg absolute counts tripled during ECP treatment while oligoclonal TCRVB families disappeared in all responsive patients. Despite the limited number of the subjects treated, the rapid clinical recovery observed in these highly compromised patients enforces the indication to employ intensive ECP courses to accelerate the clinical remission and to reduce the immunosuppression. As regards the possible use of the Treg or the TCRVB oligoclonal subsets modifications as prognostic factors during ECP treatment, the true involvement of these subpopulations is up to date, unclear. In particular Cyclosporine A might inhibit the production of Tregs and probably, other biological markers must be individuated.