OBJECTIVE Imatinib plasma trough levels (C(min)) have been reported to correlate with treatment outcomes in patients with gastrointestinal stromal tumors (GISTs). We therefore have evaluated the correlation between imatinib C(min) and the clinical characteristics of patients with GIST. METHODS Steady-state imatinib C(min) in 107 patients with GIST who were taking imatinib 300 to 800 mg/d was measured. RESULTS In patients treated with imatinib 400 (n = 92), 300 (n = 7), 600 (n = 2), or 800 (n = 11) mg/d, imatinib C(min) was 1,305 +/- 633 ng/mL, 1,452 +/- 830 ng/mL, 1,698 +/- 725 ng/mL, and 3,330 +/- 1,592 ng/mL, respectively. Of the 92 patients treated with 400 mg/d imatinib, 59 patients (63%) were men; the median age was 55 years (range, 28 to 76 years), and the median duration of imatinib use before sampling was 8.8 months (range, 0.5 to 67.6 months). The mean inter- and intrapatient variability rates were 44.7% and 26.5%, respectively. In univariate analyses, high C(min) was correlated with advanced age (P = .02), low creatinine clearance (P = .001), low hemoglobin (P = .03), and low albumin (P < .001) concentrations. Imatinib C(min) was also significantly lower in patients with (n = 18; 942 +/- 330 ng/mL) than without (n = 74; 1,393 +/- 659 ng/mL) major (ie, total or subtotal) gastrectomy (P = .002). In multivariate analysis, albumin (P = .001) concentrations, creatinine clearance (P = .002), and major gastrectomy (P = .003) were significantly correlated with C(min). CONCLUSIONS In patients with GIST, imatinib C(min) at steady state was significantly associated with albumin concentration, creatinine clearance, and previous major gastrectomy. Although its clinical impact is unclear at present time, monitoring of imatinib C(min) might be particularly important for optimal treatment with imatinib in patients who have undergone major gastrectomy.