Abstract: Osteologists often rely on single measurements, such as humeral and femoral head diameters, to estimate sex, especially when skeletons are incomplete. Measurements of 237 Bass Donated Collection skeletons provide a means of distinguishing white American females from males based on a modern sample: humeral head, female mean 42.1 mm, male mean 49.0 mm; and femoral head, female mean 42.2 mm, male mean 48.4 mm. Probabilities that bones at 1-mm increments came from females (pf) are estimated (pm = 1 − pf). An overrepresentation of one sex in the skeletons that are examined influences the probability that a bone of a certain size is from a female or male. So, probabilities are also estimated for samples consisting of an unequal number of males and females. Sample composition has its greatest effect when one sex dominates the remains that are the subject of investigation.