Abstract Interactions between tumorigenic cells and the microenvironment are increasingly recognized as integral to tumor progression in a range of human malignancies. However, the specific cellular mechanisms that are required to initiate these multistage processes are incompletely understood. Mutations in the NF1 tumor suppressor gene cause neurofibromatosis type 1, a pandemic autosomal dominant genetic disorder of the nervous system characterized by the development of neurofibromas. Neurofibromas are complex tumors composed of Schwann cells, fibroblasts, endothelial cells, and high concentrations of degranulating mast cells. Though neurofibromas are generally benign, plexiform neurofibromas can progress to malignancy. Genetic studies in cre/lox mice indicate that nullizygous loss of Nf1 in the tumorigenic Schwann cells (Krox20; Nf1flox/flox) is necessary, but not sufficient for neurofibroma formation when the microenvironment is wildtype. However, neurofibromas form with 100% penetrance in Krox20; Nf1flox/− mice that are heterozygous at Nf1 in all lineages of the tumor microenvironment (Science, 2002). Here, we addressed the role of the hematopoietic system in the tumor microenvironment by using adoptive transfer. Syngeneic Nf1+/− or wildtype (WT) bone marrow was transplanted into lethally irradiated Krox20;Nf1flox/flox mice. Krox20; Nf1flox/flox recipients transplanted with WT bone marrow (n=25) did not develop plexiform neurofibromas and had a normal lifespan. In contrast, Krox20; Nf1flox/flox mice transplanted with Nf1+/− bone marrow (n=25) consistently developed neurofibromas infiltrated with Nf1+/− mast cells. These mice had a 90% mortality at 14 months following transplantation. In complementary experiments, WT bone marrow was transplanted into irradiated Krox20; Nf1flox/− mice. Despite the remainder of the tumor microenvironment being heterozygous, WT bone marrow was sufficient to prevent tumor progression in Krox20; Nf1flox/− mice. To specifically assess the role of the mast cell compartment in tumor progression, Nf1+/− mice were intercrossed with two strains of naturally occurring W mutant mice that have variably diminished c-kit activity and mast cell function. Mice homozygous at the Wv locus have a greater than 90% reduction in c-kit activity, while W41/W41 mutants have approximately a 65–75% reduction in c-kit activity. Importantly, while Krox20;Nf1flox/flox mice transplanted with Nf1+/− bone marrow consistently develop plexiform neurofibromas, adoptive transfer of Nf1+/−; Wv/Wv or Nf1+/−; W41/W41 bone marrow cells into Krox20; Nf1flox/flox mice was sufficient to prevent neurofibroma formation. Collectively, these studies provide genetic evidence that the hematopoietic system and specifically mast cells are integral to plexiform neurofibroma formation in genetically engineered mice. These studies have therapeutic implications for NF1 since molecular therapies directed at the haploinsufficient hematopoietic cells, particularly the c-kit receptor tyrosine kinase, may have an important role in treating or preventing plexiform neurofibromas.