PURPOSE: The primary aim of this study was to determine if the humerus, the radius, and the ulna exhibit sexual dimorphism in the contemporary Croatian population, while secondary aims were to establish population-specific univariate standards and discriminant equations for the upper limb bones with known and new osteometric anatomical parameters. STUDY DESIGN: The study was conducted on the skeletal remains of the victims of the Homeland War (1991-1995). Sexual dimorphism was analyzed using standard descriptive statistics. Receiver operating characteristic curve analysis was used to determine the cut-off points for the highest accuracy in sex estimation. Standard and stepwise discriminant function analysis was performed, and discriminant function equations were established. RESULTS: All measured anatomical parameters significantly differed between males and females (P <
0.001). Univariate sex estimation accuracy using ROC was as high as 93.2 %, while multivariate sex estimation using discriminant function equations demonstrated accuracy ranging from 83.8 % to 91.2 %. CONCLUSION: This study used new osteometric anatomical parameters to establish population-specific (Croatian population) univariate and multivariate sex estimation standards and equations. Established univariate standards can be used for sex estimation when the fragmentary upper limb is discovered in a forensic context, while the established discriminant function equations can be used as the sole method or in combination with other sex estimation methods when complete upper limb bones are recovered.