Radiographs of 220 calcaneal fractures (205 patients) were reviewed retrospectively. One hundred and sixty-three fractures were intra-articular; thirty (18 per cent) of the 163 fractures were a tongue-type injury, and 133 (82 per cent) were a joint-depression injury. Plain radiographs and computerized tomography scans in the coronal and axial planes were available for 116 intra-articular fractures (106 patients). These studies were reviewed, and the 116 fractures were grouped according to the Sanders classification of calcaneal fractures and the anterior extension of the primary fracture line was evaluated. Sixty-two primary fracture lines (53 per cent) extended into one articular surface; twenty-three (20 per cent), into two articular surfaces; twenty-three (20 per cent), into a periarticular location; and eight (7 per cent), into a medial or lateral location. Sixty-seven (58 per cent) involved the calcaneocuboid joint, thirty-one (27 per cent) involved the anterior facet of the talocalcaneal joint, and ten (9 per cent) involved the middle facet of the talocalcaneal joint. A distinct anterolateral fragment was identified in 108 fractures (93 per cent). Plain radiographs failed to demonstrate the anterior extension of fifty-one (47 per cent) of the primary fracture lines. The prevalence of involvement of the anterior facet was significantly greater in Sanders type-III fractures (sixteen of thirty-five; 46 per cent) than in the other types (p < 0.01). The prevalence of involvement of the calcaneocuboid joint and the middle facet was evenly distributed among the fracture types.The primary fracture line typically extends anterior to the angle of Gissane, creating a relatively consistent anterolateral fracture fragment. The primary fracture line cannot be reliably seen on plain radiographs and is better visualized on computerized tomographic scans.