The ability of bone marrow (BM) samples to generate confluent stromal layers in long-term BM cultures (LTBMC) was used as a surrogate assay to determine the in vivo toxic effects of different conditionings on stromal cells. Here, 32 patients receiving a fludarabine-based reduced intensity conditioning regimen (FBM) were compared to those in a control group of 23 patients treated with standard busulfan/cyclophosphamide (BuCy; 14 patients) or TBI-based (TBI 12 Gy/VP16/cyclophosphamide; 9 patients) conditioning. BM was aspirated before conditioning, and at day +30 and/or at day +100, obtaining positive stromal cell growth in vitro in 58%, 47%, and 65%, respectively. FBM conditioning did not alter the ability of BM to generate stromal layers both early (day +30, 75%+) or late (day +100, 80%+) after hematopoietic cell transplantation (HCT) as compared to pre-HCT (66.6%+). FBM-treated patients formed confluent stroma significantly more often than standard-treated patients (85% vs. 38% patients; p < 0.05). In an univariate analysis, standard conditioning remained the only factor predicting stromal growth impairment after allogeneic HCT. The ex vivo-generated stromal layers from 5 female, FBM treated, sex-mismatched, and peripheral blood stem cell (PBSC) transplanted patients were analyzed by combined FISH-Y and immunofluorescence stains (Vimentin, CD14, CD45) and found to be exclusively of recipient origin. We conclude that FBM reduced intensity conditioning results in reduced, if any, stromal damage as compared to standard myeloablative treatment. The novel, donor-derived, hematopoiesis in FBM patients after allogeneic transplantation is supported and maintained by a host-derived BM stromal microenvironment.