<h3>Background:</h3> Nowadays, rheumatologists face challenges in finding an effective method to classify and treat patients with undifferentiated arthritis (UA). There is a need for new tools that could ensure accurate characterization of inflammatory processes in these patients. <h3>Objectives:</h3> To investigate if a characterization of UA patients using US may help to fulfill the 2010 ACR/EULAR RA classification criteria in a real-life cohort. <h3>Methods:</h3> We conducted a cross sectional study in two rheumatology care clinics. Patients not fulfilling the 2010 ACR/EULAR RA criteria were included. On the examination day, all patients underwent a physical examination, radiographs and US. The 7-joint US score (US 7) was adopted to scan all patients. US was performed according to EULAR criteria and interpreted by OMERACT definitions. Greyscale and power Doppler synovitis and tenosynovitis were scored. Bone erosions were also evaluated during the US examination. <h3>Results:</h3> A total of 204 patients were included. The diagnosis was modified from UA to RA in 86 (42.1%) patients. The greater proportion of synovitis detected by US was the main parameter that allowed changing the diagnosis from UA to RA, and modified the final score of the 2010 ACR/EULAR classification criteria, from a mean (±SD) of 4.6 (0.5), by clinical examination, to 6.5 (0.6) by US. The changes in the score of the 2010 ACR/EULAR classification criteria were from score 4 to score 6 in 6 (7%) patients; from 4 to 7 in 24 (27.9%) patients; from 5 to 6 in 42 (48.8%) patients; from 5 to 7 in 5 (5.8%) patients and from 5 to 8 in 5 (5.8%) patients. In addition to synovitis, a wide range of tenosynovitis and bone erosions were detected by US. Synovitis was more frequently detected in 2^ndMCPj followed by 2^ndMTPj and 5^thMTPj. The tendons of the wrist, 2^nd and 3^th finger were the most affected. In relation to bone erosions, 2^ndMCPj and 5^thMTPj where the joints with more proportion of anatomical damage. <h3>Conclusion:</h3> US demonstrated to be useful to help accurately classify as RA, patients previously diagnosed with UA. <h3>Disclosure of Interests:</h3> Marwin Gutierrez: None declared, Chiara Bertolazzi: None declared, Edwin Castillo: None declared, Denise Clavijo Cornejo: None declared, Luis Carlos Rodriguez Delgado: None declared, Jaime Mendoza Torres: None declared, Carlos Pineda: None declared, Pedro Santos-Moreno Grant/research support from: Dr Santos has received research grants from Janssen, Abbvie and UCB, Speakers bureau: Dr Santos has received speaker fees from Sanofi, Lilly, Bristol, Pfizer, Abbvie, Janssen and UCB