In vehicle collisions, the lap belt should engage the anterior superior iliac spine (ASIS) of occupants. Three-dimensional (3D) shapes of bones and soft tissues around the pelvis were acquired using a computed tomography (CT) scan of 10 male and 10 female participants wearing a lap belt in the upright and reclined postures. In the upright sitting posture, the thigh height was larger with a higher BMI while the ASIS height did not change significantly with BMI. As a result, the height of the ASIS relative to the thigh (ASIS-thigh height) became smaller as the BMI increased. Because the thigh height of females was smaller than that of males, the ASIS-thigh height was larger for females than for males. As the ASIS-thigh height was larger, the overlap of the lap belt with the ASIS increased. Thus, the lap belt overlapped more with the ASIS for the females than for the males. In the reclined sitting posture, the lap belt is positioned at the abdominal fold for the high-BMI participants while the lap belt moved to the abdominal area for the low-BMI participants. To evaluate the difference in lap belt/pelvis interaction due to body shape, impact simulations were conducted for rear seat occupants using a human finite element model THUMS. The high and low BMI model were developed by morphing the torso and the thigh shape based on the CT data. For the high BMI model, the lap belt had sufficient time to catch the pelvis due to thick soft tissue around the pelvis. As the pelvis excursion was large and the torso tilted backward, the shoulder belt path moved close to the neck. For the low BMI model, the abdominal soft tissue was thin and the lap belt force concentrated in the upper location of the pelvis. Then, the pelvis tilted rearward, and submarining occurred. The simulation results showed that low BMI occupants had a higher risk for submarining and abdominal injuries.