Sex determination is a pivotal step in forensic and bioarchaeological fields.Generally,scholars focus on metric or qualitative morphological features,but in the last few years several contributions have applied geometric-morphometric (GM) techniques to overcome limitations of traditional approaches. In this study,we explore sexual dimorphism in modern human tali from three early 20th century populations (Sassari and Bologna, Italy; NewYork, USA) at intra- and interspecific population levels using geometric morphometric (GM) methods.Statistical analyses were performed using shape,form,and size variables.Our results do not show significant differences in shape between males and females, either considering the pooled sample or the individual populations. Differences in talar morphology due to sexual dimorphism are mainly related to allometry, i.e. size-related changes of morphological traits. Discriminant function analysis using form space Principal Components and centroid size correctly classify between 87.7% and 97.2% of the individuals.The result is similar using the pooled sample the individual population, except for a diminished outcome for the New York group (from 73.9% to 78.2%). Finally, a talus from the Bologna sample (not included in the previous analysis) with known sex was selected to run a virtual resection, followed by two digital reconstructions based on the mean shape of both the pooled sample and the Bologna sample, respectively.The reconstructed talus was correctly classified with a P post between 99.9%and100%,demonstrating that GM is a valuable tool to cope with fragmentary tali,which is a common occurrence in forensic and bioarchaeological context.