Twenty pairs of cadaveric humeri were used to compare the rotational stabil
ity of proximal humeral prostheses fixed by proximal cementation with the s
tability offered by press fit or full cementation. For each proximally ceme
nted specimen, only the upper portion of the prosthesis was coated with cem
ent. For the Fully cemented specimens, a cement restrictor was used just di
stal to the prosthesis, and a finger-packing technique was used to fill the
proximal humeral medullary canal. Torque was applied to the Morse taper of
the prostheses, and rotational micromotion was measured at the level of th
e osteotomy. In each of 11 pairs of cadaveric humeri, one side was press fi
t and the contralateral side was proximally cemented; in each of 9 pairs, p
roximal cementation was compared with full cementation. Proximally cemented
prostheses' micromotion was significantly less than that of press-fit pros
theses (P =.0016). There was no difference in micromotion between proximal
cementation and full cementation (P =.82). Proximal cementation increased i
nitial fixation over press fit. Full cementation did not increase rotationa
l stability.