BACKGROUND Neurofibromatosis type 1 (NF1) is a genetic disorder that predisposes affected individuals to formation of benign neurofibromas, peripheral nerve tumors that can be associated with significant morbidity. Loss of the NF1 Ras-GAP protein causes increased Ras-GTP, and we previously found that inhibiting MEK signaling downstream of Ras can shrink established neurofibromas in a genetically engineered murine model. PROCEDURES We studied effects of MEK inhibition using 1.5 mg/kg/day PD-0325901 prior to neurofibroma onset in the Nf1 (flox/flox); Dhh-Cre mouse model. We also treated mice with established tumors at 0.5 and 1.5 mg/kg/day doses of PD-0325901. We monitored tumor volumes using MRI and volumetric measurements, and measured pharmacokinetic and pharmacodynamic endpoints. RESULTS Early administration significantly delayed neurofibroma development as compared to vehicle controls. When treatment was discontinued neurofibromas grew, but no rebound effect was observed and neurofibromas remained significantly smaller than controls. Low dose treatment of mice with PD-0325901 resulted in neurofibroma shrinkage equivalent to that observed at higher doses. Tumor cell proliferation decreased, although less than at higher doses with drug. Tumor blood vessels per area correlated with tumor shrinkage. CONCLUSIONS Neurofibroma development was not prevented by MEK inhibition, beginning at 1 month of age, but tumor size was controlled by early treatment. Moreover, treatment with PD-0325901 at very low doses may shrink neurofibromas while minimizing toxicity. These studies highlight how genetically engineered mouse models can guide clinical trial design.