To use a 3-dimensional computed tomography (CT) bone database to evaluate the risk of distal contact between the rasp tip and the endosteal cortical bone. Using a 3-dimensional CT bone database, the rasps for Exeter stems of 125 mm in length and body size 1, with a femoral offset of 37.5, 44, or 50 mm were compared with those for Exeter stems of 150 mm in length and same body size with the corresponding femoral offset. Rasp geometry was determined using an engineering drawing software. Of the 631 femurs in the database, 238 (187 Caucasian and 51 Asian) were of appropriate femoral offset and proximal body size to receive a stem with an offset of 37.5, 44, or 50 mm. Of these, 145 (115 Caucasian and 30 Asian) femurs were of champagne-flute type; the prevalence was comparable between the 2 populations (61% vs. 59%, p=0.729). When using the 150-mm rasp, 70 (55 Caucasian and 15 Asian) of the 238 femurs had distal contact between the rasp and femoral cortex; the prevalence was comparable between the 2 populations (29% vs. 29%, relative risk=1.0, p=1.0). Distal contact between the rasp and femoral cortex occurred more commonly in champagne-flute-type femurs than other femurs in the anteroposterior plane (28% [41/145] vs. 2% [2/93], relative risk=13.1, p<0.001) and in the mediolateral plane (27% [39/145] vs. 14% [13/93], relative risk=1.92, p=0.019). When using the 125-mm rasp, only one femur (with a canal flare index of 4.52) had distal contact in the mediolateral plane with an offset of 37.5 mm. Distal contact between the rasp and femoral cortex occurred more often with the 150-mm rasp than the 125-mm rasp in both planes (p<0.001). The use of a shorter stem may enhance anatomic fit in patients with a narrow femoral canal and prevent distal contact between the rasp and femoral cortex.