The aim of this study was to evaluate the reliability of the outcome measures in rheumatology (OMERACT) definitions for ultrasound (US) elementary lesions in gout through an image reading exercise. Images from patients with gout (static images and videos) were collected. As an initial step, we carried out a image reading exercise within the experts of the Pan-American League of Associations for Rheumatology (PANLAR) US Study Group (n = 16). The following step consisted in a web-based exercise with the participation of larger number of sonographers (n = 63) from different centers. Images were rated evaluating the presence/absence of any US elementary lesion. Inter- and intra-reader reliabilities were analyzed using kappa coefficients. Participants were stratified according to their level of experience. In the first exercise, inter-reader kappa values were 0.45 for aggregates, 0.57 for tophus, 0.69 for erosions, and 0.90 for double contour (DC). Intra-reader kappa values were 0.86, 0.76, 0.80, and 0.90, respectively. The web-based exercise showed inter-reader kappa values for aggregates, tophus, erosions, and DC of 0.42, 0.49, 0.69, and 0.79, respectively. The intra-reader kappa values were 0.62, 0.69, 0.77, and 0.85, respectively. Reliability was not influenced by the sonographer’s level of experience. The reliability of the new OMERACT US definitions for elementary lesions in gout ranged from moderate to excellent, depending on the type of lesion.