Galectin-3 has been suggested as a pro-inflammatory mediator in animal arthritis and rheumatoid arthritis (RA). We aimed to study the serum level of galectin-3 in patients with newly diagnosed RA and associations with disease profile, Magnetic resonance imaging (MRI) findings and seromarkers of synovial matrix inflammation. One hundred and sixty DMARD naïve patients newly diagnosed with RA were included (CIMESTRA study). Clinical, serological and imaging data were recorded before treatment and at 6 weeks, 3 and 12 months. Galectin-3 and hyaluronan (HYA) were measured by ELISA (R&D and Corgenix, USA), and the N-terminal propeptide of type III collagen (PIIINP) by radioimmunoassay (Orion Diagnostica, Finland). One hundred and nineteen, 87 and 60 blood donors served as controls for galectin-3, HYA and PIIINP, respectively. Baseline galectin-3 was significantly elevated in anti-CCP positive (4.2 μg/l IQR [3.6;6.1]) patients as compared with anti-CCP negatives (4.0 μg/l [2.6;4.9], P = 0.05) and controls (3.8 μg/l [3.0;4.8], P < 0.01). During treatment, galectin-3 remained elevated, but increased transiently with peak values at 6 weeks. Galectin-3 correlated with baseline smoking, anti-CCP, and with MRI erosion score after 1 year of follow-up. HYA and PIIINP were elevated (P < 0.001) irrespective of anti-CCP status and correlated positively with synovitis assessed clinically and by MRI. HYA and PIIINP did not correlate with galectin-3. These observations indicate that HYA and PIIINP mainly reflect expansive synovitis proliferation while galectin-3 is more closely linked to autoimmunity, smoking and joint destructive processes.