OBJECTIVES: Treat-to-target strategies for the treatment of rheumatoid arthritis (RA), as recommended in current treatment guidelines, require the use of reliable early markers of treatment response in order to adapt therapy in a timely way in patients with insufficient response. The objective of this study was to evaluate and compare the performance of different clinical indicators in predicting clinical response at one year. METHODS: This was a post-hoc analysis of data from the RAPID1 randomised, placebo-controlled Phase III study. 382 severe RA patients treated with certolizumab pegol 200mg and no prior TNF inhibitor were assessed by Week 12 (bW12) and at Week 52 (W52). “Insufficient response” was defined using the ACR response criteria, compared to baseline, DAS28 score (Δ<1.2), HAQ score (Δ<0.22) and CDAI (score<22). The ability of insufficient response to these markers bW12 to predict failure to achieve ACR50 at W52, was compared in terms of positive predictive value (PPV), specificity and sensitivity. RESULTS: At W52, 149 (38.1%) patients met the ACR50 response criterion. For all bW12 outcome indicators, the specificity and the PPV was >80%. The higher PPVs were observed for bW12 ACR20 non response (1.00 [95%CI: 0.93-1.00]), bW12 reduction in DAS28 (0.95 [95%CI: 0.82-0.99]) and bW12 CDAI score (0.93 [95%CI: 0.85-0.97]). For bW12 ACR20 non response, bW12 reduction in DAS28 and bW12 CDAI score, specificities were respectively 1.00 (95%CI: 0.97-1.00), 0.99 (95%CI: 0.94-1.00) and 0.95 (95%CI: 0.90-0.98). The highest sensitivity was observed for the bW12 ACR50 (0.70 [95%CI: 0.63-0.76]). CONCLUSIONS: All these early response markers have high PPV and high specificity but low-to-moderate sensitivity: 80% to 100% of patients identified as “insufficient-responders” by W12 would fail to fulfil the ACR50 response at W52. Such predictability data for certolizumab pegol should help physicians to make early decisions about the potential discontinuation of the treatment.