Human body models (HBMs) offer the potential to predict injury risk across broad populations. Adapting baseline HBM geometry to different body shapes, however, is often done using statistical shape models based on age, sex, height and weight, which generally explain less than half of the population variability. Here we explore that residual ribcage shape variability within a subpopulation (average height/weight males) and its influence on rib fracture risk using the SAFER HBM in simulated frontal and side impacts. Principal component (PC) analysis of rib centroidal path curves was used to define the overall ribcage geometry variation from 89 males, and the SAFER HBM was then morphed along each PC dimension. Six PCs described 90% of the geometric variance, with the first two PCs (generally representing variations in rib angles and lengths) describing 75% of this variance and having the largest influence on frontal and side impact rib fracture risk. Changing rib angles or rib lengths altered the left-right balance of seatbelt loading in the frontal impact, increasing rib fracture risk. In side impacts, rib fracture risk increased with increasing ribcage width. Ribcage shape models should consider parametrisation beyond broad demographic variables to represent variability critical for rib fracture risk.