OBJECTIVE: To estimate specificity, sensitivity, positive predictive value and negative predictive value of tests and signs used for the diagnosis of cranial cruciate ligament failure in dogs. METHODS: One stifle in each of 42 dogs was examined: 25 "affected" and 17 "control" dogs. All dogs were subjected to the following tests when conscious: cranial drawer, tibial compression, patellar tendon palpation and palpation of the medial aspect of the joint. Under general anaesthesia, cranial drawer and tibial compression tests were repeated and a lateral stifle radiograph was taken to evaluate changes of the infrapatellar fat pad. The results were analysed using a 2 x 2 table method. Sensitivity, specificity, positive predictive value and negative predictive value were estimated. RESULTS: The sensitivity of the cranial drawer and tibial compression tests was surprisingly low when performed on conscious patients but significantly better when performed under anaesthesia. Similarly, palpation of the medial aspect of the stifle joint cannot be considered a reliable indicator of cranial cruciate ligament injury. Patellar palpation and radiographic assessment showed excellent sensitivity, specificity, positive predictive value and negative predictive value. CLINICAL SIGNIFICANCE: In the diagnosis of cranial cruciate ligament, it is essential that the clinician is aware of each test's features and limitations to reduce the risk of misdiagnosis.