Objective: To examine the clinical features associated with an ultrasound (US) diagnosis of synovitis and/or erosions in patients suspected of inflammatory arthritis, and the factors associated with this evolution in patients with a normal initial US. Patients and methods: Cross-sectional: the records of 144 patients who underwent US for suspected inflammatory arthropathy were categorized into synovitis and/or erosions present or not. Longitudinal: of 58 patients without synovitis and/or erosions, 30 could be located and 19 agreed to be studied (two were asymptomatic and refused, nine could not be reached). Analyses: univariable descriptive analyses were performed. Age, gender, variables significant (p<0.05) in the univariable analyses, and those clinically relevant were examined by logistic regression for the cross-sectional study. The metric properties of US compared to overall clinical assessment were also examined. Results: Age, gender, ethnicity and symptoms'duration were comparable in patients with and without synovitis and/or erosions. Wrist swelling (history) and the number of swollen wrist/hand joints were associated with synovitis and/or erosions by US; morning stiffness, sicca symptoms and low back pain were negatively associated with synovitis and/or erosions. Four patients evolved into an inflammatory arthropathy but no features distinguished them from those who did not evolve into an inflammatory arthropathy. The sensitivity, specificity, and overall accuracy of US, compared to the clinical assessment were 98.9%, 94.1% and 98.1%, respectively. Conclusions: US is an adequate tool for the assessment of inflammatory arthropathy; however, patients with a single negative US at initial clinical presentation still need to be followed for the eventual development of an overt arthropathy. © Copyright Clinical and Experimental Rheumatology 2008.