Injury risk curves for the skeletal knee–thigh–hip (KTH) relate peak force applied to the anterior aspect of the flexed knee, the primary source of KTH injury in frontal motor-vehicle crashes, to the probability of skeletal KTH injury. Previous KTH injury risk curves have been developed from analyses of peak knee-impact force data from studies where knees of whole cadavers were impacted. However, these risk curves either neglect the effects of occupant gender, stature, and mass on KTH fracture force, or account for them using scaling factors derived from dimensional analysis without empirical support. A large amount of experimental data on the knee-impact forces associated with KTH fracture are now available, making it possible to estimate the effects of subject characteristics on skeletal KTH injury risk by statistically analyzing empirical data. Eleven studies were identified in the biomechanical literature in which the flexed knees of whole cadavers were impacted. From these, peak knee-impact force data and the associated subject characteristics were reanalyzed using survival analysis with a lognormal distribution. Results of this analysis indicate that the relationship between peak knee-impact force and the probability of KTH fracture is a function of age, total body mass, and whether the surface that loads the knee is rigid. Comparisons between injury risk curves for the midsize adult male and small adult female crash test dummies defined in previous studies and new risk curves for these sizes of occupants developed in this study suggest that previous injury risk curves generally overestimate the likelihood of KTH fracture at a given peak knee-impact force. Future work should focus on defining the relationships between impact force at the human knee and peak axial compressive forces measured by load cells in the crash test dummy KTH complex so that these new risk curves can be used with ATDs.