Joint effusion is one of the classic radiographic signs of osteoarthritis, but no information is available regarding clinicians' ability to detect joint effusion radiographically. This study determined the accuracy and precision of experienced and inexperienced observers in detecting joint effusion of the distal interphalangeal (DIP), metacarpophalangeal (MCP) and midcarpal joints on radiographs. Fresh cadaverous forelimbs were loaded in a material testing machine to mimic a standing horse. The joints were gradually distended and lateromedial and dorsopalmar radiographs were acquired. The images were assessed by three experienced and three inexperienced observers. This study showed that the sensitivity was high for both groups for all joints and projections. Specificity was high for the experienced group but low for the inexperienced group. There was a significant positive correlation between joint distension and severity of joint effusion when reported by experienced observersfor most views (except DIP joint), but only for the MCP joint for inexperienced observers. Interoperator agreement was poor to fair for the experienced group and poor for the inexperienced group. In conclusion, the accuracy and precision of grading joint effusion on radiographs was highly subjective and related to experience.