The objective of this study was to determine whether current design validation methods for Vibration Isolation & Stabilization (VIS) systems can use exercise data collected from stable surfaces for stability requirements. The VIS must attenuate exercise-induced accelerations transmitted to the spacecraft structure and payloads to below microgravity allocations. Prior ground-based tests collected exercise data from stable surfaces, but fundamental changes in VIS design have necessitated a deeper understanding of exercises on unstable surfaces. Therefore, this study aimed to perturb subjects' center of gravity (CG) and observe the effects on exercise performance and the VIS simulation used for design validation. Fourteen subjects with prior resistive exercise experience participated in this counterbalanced, repeated measures study. Subjects performed exercise trials on both a stable floor and an unstable platform; the assigned group determined the order of the stable and unstable conditions. The subjects used a cablebased resistive exercise device (Miniature Exercise Device, MED-2) to perform three sets of eight repetitions for bent over rows, front squats, and overhead presses. The exercise performance was altered on the unstable surface compared to the stable one. The exercise motion range, or the range of the upper and lower exercise limits, increased from the stable to unstable surface on average across groups by 9.73%, 7.12%, and 6.93% for bent over rows, front squats, and overhead presses, respectively. The subjects' repetition times increased during front squats for 74.60% of repetitions across sets. The area of the CG trajectory significantly increased from the stable to unstable condition for front squats and overhead presses (p < 0.001). The order of the stable and unstable conditions did not appear to affect most metrics, except for the unstable overhead presses and the perceived exertion rating. Subjects that started with the stable condition substantially increased their perceived exertion rating for bent over rows during the unstable condition. Results from the VIS simulation analysis indicated that microgravity accelerations were greater for the unstable condition at frequencies less than 0.3 Hz. However, these accelerations did not exceed microgravity allocations.