Methods of stem fixation are a controversial aspect of revision TKA. We sought to determine which technique was superior by reviewing 475 revision TKAs done between 1986 and 2000. Of these 475 revisions, 286 major component revisions were done using 484 extended stems for fixation. Patients who died, patients who had less than 2 years follow up, or patients who had diaphyseal engaging stems were excluded from the study. The final data set included 113 revision TKAs with 202 metaphyseal engaging stems. Of the 202 stems, 107 were cemented whereas 95 were press-fit metaphyseal engaging stems. One hundred one of these were femoral stems and 101 were tibial stems. Using a modified Knee Society radiographic scoring system, 100 (93%) of the 107 implants with cemented stems were considered stable, seven (7%) were categorized as possibly loose requiring close followup, and none were loose. Of the 95 implants placed with cementless stems, only 67 (71%) were categorized as stable. Eighteen (19%) were possibly loose requiring close followup and 10 (10%) were loose (two tibial and eight femoral implants). We currently would urge caution in using cementless metaphyseal engaging stems for fixation in revision TKA.