Sex estimation is the grounds for an accurate identification of unknown human skeletal elements. The need for reliable methods distinguishing males from females based upon various skeletal elements is evident in cases of commingled, eroded and/or missing remains. The aim of this work lays on establishing criteria for sex estimation from the scapula and the clavicle in modern Greeks. A total of 147 left scapulae and 147 clavicles (66 females and 81 males) were used in the study. Eight and six measurements were taken on the scapula and clavicle respectively and data were subjected to principal components analysis (PCA) and discriminant function analysis (DFA). Posterior probabilities for the classification of each individual are also calculated. Statistical analysis was carried out using the software PAST (Paleontological Statistics) and SPSS (Statistical Package for Social Sciences) 18. The results supported the existence of pronounced sexual dimorphism, which was mainly attributed to size differences among the two groups. Univariate and multivariate methods of statistical classification showed high accuracy for all scapular and most clavicular measurements, verifying their value as sex indicators in the under study population. We recommend the use of this method for sex assessment from the scapula and the clavicle in cases exhibiting over 95% probability of correct classification. This is regardless of the overall high degree of accuracy reported here, as the method of choice in forensic contexts should always be case-driven.