Purpose: We identified the best predictors of clinical response to simple add-on tolvaptan (TLV) therapy in patients with heart failure (HF). Methods: We retrospectively enrolled 60 HF patients with excess fluid retention despite receiving adequate medical therapy including oral diuretics. All patients received simple add-on TLV (median of 7.5mg/day). They underwent right heart catheterization at baseline and after 7-day treatment. Results: Although the majority of patients were successfully treated with simple add-on TLV therapy (Group 1), but 22% (Group 2) were defined as being unsuccessfully treated because 1) HF symptom score worsened or 2) HF symptom score >6, and mean pulmonary capillary wedge pressure (PCWP) >18 mmHg and mean right atrial pressure (RAP) >10 mmHg after TLV therapy. Although the degrees of body weight reduction and increment of urine volume were similar between the two groups, HF symptom score and mean PCWP and RAP improved, and plasma BNP level decreased only in the group 1 after TLV therapy. The group 2 had lower urine sodium/creatinine ratio (UNa/UCr) and higher plasma BNP level than responders at baseline, and they were identified as the independent predictors of unsuccessful TLV therapy by multivariate logistic regression analysis. Receiver operating characteristic curve analysis showed that UNa/UCr was the strongest predictor of unsuccessful TLV therapy with cut-off point of 46.5 mEq/gCr (AUC 0.847, 95% CI; 0.718-0.976, sensitivity 77%, specificity 81%, P<0.001). Furthermore, combination with UNa/UCr 778 pg/ml best predicted unsuccessful TLV therapy with sensitivity of 54%, specificity of 100%, positive predictive value of 100%, negative predictive value of 89%, and accuracy of 90%. Conclusion: Simple add-on tolvaptan therapy ameliorated HF symptoms and provided hemodynamic improvement in majority of patients with HF, and the measurements of UNa/UCr and BNP level can help clinicians tailor HF treatment.