Glenohumeral translation and rotation were measured in 6 grossly normal, Fr
esh frozen shoulder preparations while a manual load was applied to the hum
erus. The same tests (maximum elevation, total rotation, anterior/posterior
(A/P) translation, and inferior translation) were repeated for each should
er through 8 series: 1 with the shoulder intact, 1 with the shoulder vented
, and 6 with progressively larger humeral head components after hemiarthrop
lasty. There was an inverse linear relation between humeral head component
size and all 4 outcome variables. Replacing the native head with a componen
t of equal diameter reduced elevation 20%, rotation 40%, A/P translation 50
%, and inferior translation 60% in the vented shoulder Replacing the native
head with a component of equal effective volume decreased elevation 8%, ro
tation 20%, A/P translation 25%, and inferior translation 40% in the vented
shoulder Increasing humeral head component size decreased rotation, A/P tr
anslation, and inferior translation by similar percentages and elevation so
mewhat less. Humeral head component size is better described in terms of vo
lume than in terms of diameter or offset.