Background/Purpose: We have previously (2009ACR, 2010ACR) reported that the efficacy of biologics, infliximab and adalimumab can be predictable using baseline blood a disintegrin and metalloproteinase with thrombospondin motifs 5 (ADAMTS5) mRNA level. In this study presented here, we investigated whether the efficacy of tocilizumab (TCZ) for the treatment of rheumatoid arthritis (RA) can be predicted by the baseline blood ADAMTS5 mRNA level because recently IL-6 has been reported to suppress ADAMTS5 expression. Methods: Fifty-four randomly selected active RA patients were treated with TCZ. Peripheral blood samples were collected at baseline and ADAMTS5 mRNA was quantified using real-time PCR (BiologicMate®). Baseline IL-6 mRNA was also estimated using real-time PCR. Results: Baseline ADAMTS5 mRNA levels (�L 10^-4) in the responder (2.85 ± 2.37 Index) was significantly (p=0.049) higher than that in the non-responder (1.56 ± 0.80 Index) at 12 wks‘ treatment with TCZ. Area under curve (AUC) of the ROC curve for predicting the clinical remission (DAS28<2.6) using the baseline ADAMTS5 mRNA at 12 wks was 0.75, and the cut-off value of ADAMTS5 mRNA was calculated as 1.70 Index. There observed no difference in the general patient backgrounds (average age, disease duration, dose of MTX, dose of steroid, baseline DAS28, etc) between the High-ADAMTS5 (³ 1.70 Index) group and the Low-ADAMTS5 group. However, DAS28 at 12 wks after treatment was significantly (p=0.0254) lower in the High-ADAMTS5 group (3.09 ± 1.29) than in the Low-ADAMTS5 group (3.88 ± 1.08) (Fig.1). The accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the baseline High-ADAMTS5 (³1.7 Index) for predicting the clinical remission at 12 wks with TCZ was 75.0%, 94.1%, 48.7%, 45.7%, and 94.7%, respectively. Interestingly, we observed negative correlation between baseline IL-6 and ADAMTS5 mRNA expression. Conclusion: The baseline ADAMTS5 mRNA level, which might be related to baseline IL-6, is a candidate biomarker for prediction of the response to TCZ in RA patients.