Relative Constant-Murley shoulder scores are based on an age- and sex-matched normal population from Constant's original research. The purpose of this study was to determine normal Constant scores in a contemporary population and compare these values with those originally established by Constant. Two subject groups were analyzed. The first group comprised 1620 clinic patients whose normal shoulders were scored by resident physicians instructed on the Constant scoring technique. The second group comprised 115 healthy volunteers whose normal shoulders were scored by 1 experienced physician-researcher. There were no significant score differences between the two groups, indicating that, after instruction, scoring did not depend on tester experience. In both groups, there were significant score differences between sex and age groups. Similar to Constant's study, age-related declines in scores and strength existed for both sexes but were less pronounced. Using Constant's original values to calculate relative scores can overestimate shoulder function in women aged over 40 years and men aged over 60 years. If relative Constant scores are used, absolute scores should be concurrently reported to allow comparisons with different populations.