Aim of the study: Here, we investigated the risk factors for decreased teicoplanin plasma trough concentrations relative to the initial dosing in critically ill patients. Patients and methods: Data obtained from 80 eligible critically ill patients who received intravenous teicoplanin were retrospectively analyzed. Risk factors for decreases in teicoplanin trough concentrations 72 h after administration of teicoplanin of more than 30% relative to predicted concentrations based on initial dosing setting were identified by logistic regression analysis. Results: Although prediction trough concentration and total dose of two days no significant differences were seen between the variation group and the non-variation group, actual trough concentration was significantly different between two groups (19.9±5.6 μg/ml vs 10.3±2.2 μg/ml, p < 0.001). In multivariate analysis, serum albumin ≤ 2.2 mg/dl (odds ratio [OR] = 3.003, 95% CI 1.072-8.408; p = 0.036) and SOFA score ≥ 9 (OR = 3.498, 95% CI 1.171-10.450; p = 0.025) were significant risk factors for decreased teicoplanin plasma trough concentrations. Conclusion: In critically ill patients, high SOFA score and low serum albumin were risk factors for decreased teicoplanin plasma trough concentration during initial dosing.