'Skiving' is commonly used to refer to the condition when the subchondral plate is disrupted and the overlying cartilage physically displaced without the screw tip entering the joint. In this study we sought to define radiographic parameters of skiving and compare radiographs with computed tomography (CT) for accuracy in determining joint skiving. Cadaveric specimens of the distal radius were implanted with a volar plate and screws. Arthrotomies were performed to definitively assess the positions of the screws. Standard and anatomic tilt radiographs as well as CT were performed. Orthopaedic surgeons and radiologists evaluated the images and reported whether screw penetration or skiving had occurred. For screws which penetrated or skived, measurements were made to record the distances from the screw tips to the subchondral plate. Sensitivity, specificity and percent correct interpretations were 53%, 83%, 60% respectively for radiographs; and 100%, 72%, 69% for CT. Screws penetrating the articular surface protruded an average 2.3 mm (range 2-2.6 mm) from the subchondral plate and those skiving protruded 1.4 mm (range 1-1.8 mm). This study shows that articular skiving can occur with penetration of the subchondral plate of up to 1.8 mm. CT has a greater sensitivity and lower specificity in determining skiving compared to radiographs.