BACKGROUND: Recent clinical studies have reported that patients with higher body mass index (BMI) are more likely to experience premature failure of total knee arthroplasty (TKA), lower knee scores, and perhaps more pain in the prosthetic joint. However, it is not known whether certain implant design features such as tibial stems might be associated with differences in the frequency of tibial pain in patients with higher BMIs. QUESTIONS/PURPOSES: Therefore, it was our aim to compare (1) function and pain (as measured by the New Knee Society Score, Knee Injury and Osteoarthritis Outcome Score [KOOS], and visual analog pain scores); (2) quality of life (as measured by SF-12); and (3) mechanical complications and premature revision (defined as revision before 2 years) between patients with obesity undergoing TKA (BMI \textgreater 30 kg/m(2)) who received either a stemmed or an unstemmed tibial component. METHODS: In this randomized controlled trial, 120 patients with a BMI \textgreater 30 kg/m(2) scheduled for primary arthritis TKAs and end-stage knee osteoarthritis were included. Patients were stratified into groups defined as moderately obese (BMI 30-35 kg/m(2), N = 60) and severely obese (BMI \textgreater 35 kg/m(2), N = 60) groups. Patients in each stratified subgroup then were randomized to receive either a stemmed (10 mm/100 mm) proximally cemented tibial component or the other, a standard cemented component. Patients were evaluated preoperatively and 2 years after surgery using the new Knee Society Score (KSS), KOOS, SF-12 score, and a visual analog pain score after 100 meters of walking. Although no minimum clinically important differences (MCIDs) have yet been defined for the new KSS, we considered differences smaller than 10 points to be unlikely to be clinically important; the MCID for the KOOS is estimated at 8 to 10 points, the SF-12 to be 4 points, and the visual analog scale to be 2 cm on a 10-cm scale. Patients were followed until death, revision, or for a minimum of 2 years (mean, 3 ± 0.8 years; range, 2-4 years). No patient was lost to followup before 2 years. RESULTS: Although we found that patients treated with stemmed TKAs had higher functional outcomes, the differences were small and unlikely to be clinically important (subjective KSS mean 69 ± 7 points versus 75 ± 7, mean difference 6 points, 95% confidence interval [CI] 2-11, p = 0.03; objective KSS mean 80 ± 6 points versus 85 ± 6 points, mean difference 5 points, 95% CI 0-9, p = 0.01). Compared with patients with a stemmed TKA, patients with a standard implant reported lower KOOS pain subscores (81 ± 9 versus 76 ± 8; p = 0.04) and lower KOOS symptom subscores (74 ± 7 versus 68 ± 7; p = 0.03). The proportions of patients experiencing complications were not different with the numbers available for all groups and subgroups. CONCLUSIONS: Although we detected differences in some patient-reported outcomes scores for pain and function favoring implants with stems, the differences were small and unlikely to be clinically important. Because these stems may have disadvantages, perhaps including difficulty of revision, we cannot draw a strong conclusion in support of their use. LEVEL OF EVIDENCE: Level I, therapeutic study.