BACKGROUND: Prior work shows caloric restriction (CR) can improve physical function among older adults living with obesity. However, the contribution of starting weight and inflammatory burden to CR-associated functional improvements is unclear. The primary purpose of this study was to determine if CR-associated gait speed change varied by body mass index (BMI) and plasma interleukin 6 (IL-6) at baseline and secondarily to determine the contribution of BMI change and IL-6 change to gait speed change. DESIGN, SETTING, PARTICIPANTS: Data from eight randomized control trials were pooled, with 1184 participants randomized to CR (n=661) and No CR (n=523) conditions. All studies assessed outcomes before and five or six months after assignment to CR or No CR. MEASUREMENTS: BMI and IL-6 were assessed at baseline using standard procedures. Gait speed was assessed with the six-minute walk test or 400m walk test. Baseline BMI/IL-6 subgroups were constructed using BMI≥35 kg/m2 and IL-6>2.5 pg/mL thresholds. Participants with BMI≥35 kg/m2 were grouped into class 2+ obesity and BMI<35 kg/m2 into class 1- obesity; IL-6>2.5 pg/mL were grouped into high IL-6, and <2.5 pg/mL as low IL-6 (class 2+ obesity/high IL-6: n=288, class 2+ obesity/low IL-6: n=143, class 1- obesity/high IL-6: n=354, or class 1- obesity/low IL-6: n=399). All analyses used adjusted general linear models. RESULTS: Gait speed significantly improved with CR versus non-CR [mean difference: +0.02 m/s (95% CI: 0.01, 0.04)]. CR assignment significantly interacted with BMI/IL-6 subgroup membership (p=0.03). Greatest gait speed improvement was observed in the class 2+ obesity/high IL-6 subgroup [+0.07 m/s (0.03, 0.10)]. No other subgroups observed significant gait speed change. For each unit decrease in BMI, gait speed change increased by +0.02 m/s (p<0.001; R2=0.26), while log IL-6 change did not significantly affect gait speed change [+0.01 m/s (p=0.20)]. CONCLUSIONS: Only the class 2+ obesity/high IL-6 subgroup significantly improved gait speed in response to CR. Improvement in gait speed in this subgroup was driven by a larger decrease in BMI, but not IL-6, in response to CR. Individuals with class 2+ obesity and high IL-6 are most likely to show improved gait speed in response to CR, with improvement predominantly driven by reductions in BMI.