Background/Purpose: The potential of the 2010 ACR/EULAR rheumatoid arthritis (RA) classification criteria to discriminate early arthritis patients according to their clinical and radiological outcomes needs to be confirmed. Methods: Consecutive patients with at least 3 swollen joints (SJC≥ 3) were recruited; duration of disease was more than 1 and less than 12 months; microcrystalline arthritides and connective tissue diseases were actively excluded. All patients were treated with the target of SJC= 0, using whatever DMARDs was required. According to the 2010 RA criteria, our cohort was classified in two groups: RA and non-specific inflammatory arthritis (NSIA). Both groups were compared at baseline and at 18 months into disease according to radiological and clinical outcomes: Sharp erosion score (classified erosive if ≥ 3), Health Assessment Questionnaire (HAQ), Disease Activity Score in 28 joints (DAS28-CRP), and pain (0-100 mm VAS). Remission rates at 18 months were calculated based on DAS 28 and ACR 2011 remission criteria. Results: A total of 422 patients were available at baseline. Of these, 319 (75.6%) were classified as RA. Based on clinical outcome measures, RA patients had more severe disease than NSA patients at baseline (higher DAS 28, HAQ and pain scores measures – p<0.001). At 18 months, this initial clinical discrepancy had now disappeared, but more RA patients had progressed to erosive status than NSIA patients (54% vs 33%; p<0.001). Using two different definitions of remission, there was no difference in remission rates. Conclusion: In this early arthritis cohort actively treated to SJC= 0, there was no difference in the 18-month clinical outcomes (HAQ, DAS 28, Pain, Remission rates) between 2010 ACR/EULAR criteria-defined RA and NSIA patients. Patients with RA had a worse radiological outcome, but significant joint damage occurred in one third of NSIA patients. Early intensive treatment of NSIA patients, and not only of early RA patients, thus appears warranted.