A pharmacokinetic study in patients with gastrointestinal stromal tumors (GIST) suggested that imatinib plasma concentration may decrease following long-term exposure. We assessed changes in imatinib plasma trough levels (C(min)) during long-term treatment. Follow-up (FU) imatinib C(min) was measured in 65 patients who received the same dose of imatinib for at least 9 months after previous (initial) tests. After exclusion of 7 patients who had been treated with imatinib for over 2 years at the time of initial testing, 58 patients were included in this analysis. The median intervals from initiation of imatinib to initial testing and from initial to FU testing were 5.5 months (range, 0.5-24.0 months) and 13.0 months (range, 9.6-17.9 months), respectively. Mean inter- and intra-subject variability values were 47.7% and 20.9%, respectively, at initial measurements, and 45.2% and 19.4%, respectively, at FU. Mean FU imatinib C(min) (1,370 ± 661 ng/mL) was significantly higher than mean initial C(min) (1,171 ± 573 ng/mL; p = 0.003). Compared with initial C(min), FU C(min) was decreased in 22 patients and increased in 36, with median changes of 13% and 32%, respectively. Multivariate analysis showed a significant correlation between the ratio of FU to initial imatinib C(min) and that of albumin (r = -0.39, p = 0.003). During long-term treatment, imatinib C(min) did not decrease significantly but remained stable or increased in most patients. Changes in imatinib C(min) were associated with changes in albumin concentration. Monitoring of imatinib C(min) only for concerns about time-dependent increases in imatinib clearance is not necessary.