We performed a systematic literature review on the reliability, criterion and construct validity, discriminant validity, contribution in regard to clinical examination and predictive validity of ultrasonographic (US) synovitis in rheumatoid arthritis (RA). Medline, Embase, Cochrane Library, and abstracts presented in meetings and hand search articles dealing with these parameters were collected. 190 studies with a total of 10 383 patients were analyzed. Inter and intra-observer reliability were good to excellent. Comparison to histological and macroscopic evaluation revealed overall good correlation. There was a concordance between Magnetic Resonance Imaging synovitis and US synovitis. US is more sensitive than clinical examination in detection of synovitis. US can differentiate healthy subjects from those with RA, nevertheless it could be some synovial hypertrophy in healthy subjects. There were some differences between RA and other inflammatory arthridies, without diagnostic confidence. US synovitis is reproducible, a "real" synovitis, sensitive and discriminatory, supporting the place of routine use of sonographic evaluation of patients with rheumatoid arthritis.