This paper tests the validity of the circumference at the nutrient foramen of the tibia for sex determination on two human osteological collections: the Lisbon Collection of Identified Skeletons (n = 160) and the São Martinho medieval archaeological collection from Leiria, Portugal (n = 57). The methodological approach chosen to conduct the test was the sectioning point method. The overall accuracy (males and females combined) in the Lisbon Collection was 78% and in the São Martinho Collection it reached 90%. In both series, females were more accurately classified than males. This paper advocates the importance of developing and using sample specific methods for sex determination in modern and in archaeological human osteological assemblages. Copyright © 2010 John Wiley & Sons, Ltd.