In forensic anthropology, methods for estimating sex from the skeleton have historically been developed on skeletal collections comprised exclusively U.S. Black and white populations and thus inherently rely on ancestry estimation as a foundational component for the construction of the biological profile. However, these population-specific methods of sex estimation fundamentally limit the applicability, accuracy, and reliability of the method for use on remains of diverse population affiliations (ancestry). A reliable and population-inclusive method for estimating sex in the skeleton can serve as a useful tool for forensic investigators, especially in cases with unidentified remains where the population affiliation is indeterminate and in light of recent discourse involving the removal of ancestry estimation from the biological profile as a whole This study assessed the reliability of population-inclusive nonmetric and metric sex estimation from 3D-volume rendered computer tomography (CT) scans of the skull. The study sample was comprised of 431 individuals (242 males and 189 female) from the New Mexico Decedent Imaging Database (NMDID) and included a relatively equal distribution of African American, Asian American, European American, Latin American, and Native American population affinities. The images were obtained from the CT slices using 3D-reconstructions and volume rendering technique (VRT) in the Digital Imaging and Communications in Medicine (DICOM) viewer, exported to Meshmixer™ and then processed to isolate the skull from the postcranial skeleton and remove identifying objects. In Meshmixer™, nonmetric traits were scored following Buikstra and Ubelaker (1994) and Walker (2008) and included the supraorbital ridge/glabella, supraorbital margin, mastoid process, mental eminence and nuchal crest. The metric traits, following Spradley and Jantz (2011), included 18 points of measurement of the cranium and mandible. Binary logistic regression (BLR) and discriminant function analyses (DFA) were used to produce models and probabilities from the nonmetric a metric data respectively and an additional binary logistic regression was developed that combined both the nonmetric and metric data. Overall, the population-inclusive nonmetric and metric model produced classification accuracies that ranged from 81-87% and 86.7-87% respectively, and performed as well as population-specific models in estimating sex and were not significantly different from population-specific accuracies. When some of the population-specific models were applied across population, particularly the European American model, the classification accuracy was significantly reduced relative to the population-inclusive model. Intraobserver error was assessed for the nonmetric and metric data collection and confirmed that the nonmetric and metric methods of data collection for the volume-rendered images was consistent. The results of this study indicate that a population-inclusive nonmetric and metric models of sex estimation using the skull can be used in place of more traditional population-specific models in cases where ancestry is unknown, indeterminate, or in the event ancestry is removed from the biological profile.