Gastrointestinal stromal tumor (GIST) is the most common sarcoma arising in the gastrointestinal (GI) tract. Imatinib mesylate (imatinib) is efficacious in treating advanced and metastatic GIST. Patients undergoing resection of GIST realize a highly variable median disease-free survival (DFS). In the absence of prospective data, we conducted a randomized, phase II study to assess the safety and efficacy of preoperative and postoperative imatinib for the treatment of GIST. Nineteen GIST patients undergoing surgical resection were randomized to receive 3, 5, or 7 days of preoperative imatinib (600 mg daily). Patients received postoperative imatinib for 2 years. Perioperative adverse events were compared with those in an imatinib-naïve historical control. The efficacy of imatinib was assessed by (18)fluorodeoxyglucose positron emission tomography ((18)FDG-PET), dynamic computed tomography (dCT), terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay, and DFS. Imatinib did not affect surgical morbidity as compared with an imatinib-naïve cohort (p >/= 0.1). Most patients responded to preoperative imatinib by (18)FDG-PET and dCT (69% and 71%, respectively). Tumor cell apoptosis increased by an average of 12% (range 0-33%) and correlated with the duration of preoperative imatinib (p = 0.04). Median DFS of patients treated with surgery and imatinib was 46 months (range 10-46 months). Tumor size was a predictor of recurrence after postoperative imatinib (p = 0.02). Imatinib appears to be safe and may be considered for patients undergoing surgical resection of their GIST. Radiographic response and tumor cell apoptosis occur within the first week of imatinib therapy.