105 patients with low back pain (LBP) were randomized to receive treatment with either lumbar stabilization exercises (LSE) or manual therapy (MT). A clinical prediction rule (CPR) predicting successful outcomes based on CPR status following LSE was tested for validity. After 8 weeks, the present study was unable to validate CPR, as the interaction between treatment and CPR status was absent, despite pairwise comparisons suggesting patients with positive statuses to experience less disability following LSE. A modified CPR, containing only 2 of the 4 CPR variables, was able to predict successful outcomes, but was derived retrospectively, and as such, will require more research.