Several studies have attempted to describe the morphology of the hyoid bone, while other authors have focused on discovering the role of this bone in the occurrence of fractures. Hyoid fractures are known to be dependent on the force applied against the bone, or on the location at which the force is applied. We wished to assess the value of defining one or more models of the hyoid bone by analyzing variations in the size and angle of the various component parts of the bone relative to the sex and morphology of an individual (height and weight) in a sample of 72 bones obtained during forensic autopsy at our institution. Statistical analyses were developed using SAS software (Statistical Analysis System, version 9.2). We observed that the length of the hyoid bone and the angle between the greater horns differed significantly between men and women. Length was significantly greater in men (38.20 +/- 4.67 mm) than in women (30.49 +/- 7.90 mm) and the angle between the greater horns of the hyoid bone was larger in women (36.46 +/- 13.77 degrees) than in men (27.56 +/- 13.02 degrees). There was also a statistically significant correlation between the body mass index of an individual and the length of the hyoid bone. As weight increased, the hyoid bone was found to be longer. The weight of an individual was also significantly correlated with the angle of the hyoid bone, with lower weight resulting in larger angles of the bone. Furthermore, hierarchical classification enabled the hyoid bone to be differentiated into two groups or clusters according to anthropometric measurements. ROC curves were used to determine threshold values of length, width and angle to classify the hyoid bones in these two clusters: the first was composed of individuals with longer hyoid bones, and the second of individuals with greater hyoid bone widths and wider angles. Logistic regression showed male gender was more frequently associated with the first group. The morphology of the hyoid bone can be differentiated according to the gender and corpulence of an individual because these parameters are correlated. These findings are crucial in establishing a protocol for modelling the mechanism of fracture of the hyoid bone in strangulation. Two models of the hyoid bone appear to be needed to meet the practical requirements that are the purpose of these biomechanical studies.