ObjectivesThis retrospective study evaluates the dog-related factors of variation influencing the outcome of the flexion test (FT), when performed to localize pain to a joint area, on a large group of canine orthopedic patients.Materials and methodsThe selection criteria for this retrospective study were dogs undergoing a FT in a referral orthopedic clinic between 2009 and 2020 with a complete medical record. The canine FT, described in a previously published protocol, was performed on dogs presented with an orthopedic problem. In summary, a dog's joint, identified as suspected of an orthopedic problem according to the clinical examination, was flexed for 1 min before walking 15 m on a hard and even surface. The FT was considered positive if the lameness increased after the application of the FT and negative when it remained unchanged. Statistical analysis was performed to determine which of the following criteria could influence the outcome of the flexion test: age, gender, neutered status, weight category, tested joint and initial lameness score.ResultsOver 1,161 patients' files were collected and analyzed for this research. The FT showed 82.8% (95%IC: 80.5–84.9) of true positives and 17.2% of false negatives. None of the patient's intrinsic characteristics influenced the outcome of the test (age, gender, neutered status, and weight category). The orthopedic parameters, such as the initial lameness score and the tested joint, showed to have a statistically significant influence on the outcome of the test.Clinical relevanceThe FT is an easy-to-perform technique presenting reliable results on most joints. This test presents an interest when performed in addition to a complete orthopedic examination to localize pain to a joint area. Only the orthopedic pathological-related parameters such as the lameness score and the tested joint seem to influence the outcome of the FT. The FT is not influenced by the physiological-related characteristic of the patient (age, weight category, sex, and neutered status).