This study evaluates the between-day reliability of a newly developed trunk perturbation test and compares mechanical response during known and unknown conditions. Mechanical trunk responses were measured in 17 female subjects during unloading and loading perturbations of the abdomen (A: preloaded abdomen condition) and low back (B: preloaded back condition). The loading perturbation increased the preload from 5.5 kg to a 10.9 kg pull on the trunk whereas the unloading perturbation decreased the pull from 5.5 kg to 0.1 kg. A sequence of loading (known), unloading (known), and randomized loading/unloading (unknown) perturbations were performed for A and B. Between-day reliability of stopping time, trunk displacement, and velocity was quantified using intraclass correlation coefficients (ICCs). ICCs were good to excellent for all loading and unloading measures during the known (0.70-0.98) and unknown (0.64-0.94) perturbations of A and B. In general, larger trunk displacements were seen after the unknown perturbations compared with the known perturbation. The method may be used as a diagnostic tool for screening workers who are in risk of future work-related low back injuries.