Objectives: To investigate the feasibility and effects of a sensorimotor stabilization exercise intervention with and without behavioral treatment in nonspecific low back pain. Design: A three-armed multicenter randomized controlled trial. Setting: Five study sites across Germany (3 orthopedic university outpatient clinics, 1 university sports medicine department, and 1 clinical institution). Participants: Six hundred and sixty-two volunteers (N=662) (59% females, age 39±13y) with low back pain. Interventions: Sensorimotor training (SMT), sensorimotor training with behavioral therapy (SMT+BT), and usual care group (UCG; continuation of the already ongoing individual treatment regime). Intervention groups performed a 12-week (3wk center-based, 9wk home-based) program. Main Outcome Measures: Adherence, dropout rates, adverse events, and intervention effects on pain intensity, disability, and trunk torque (gain scores, repeated measures analysis of variance, α-level<0.05). Results: In total, 220 participants received SMT, 222 received SMT+BT, and 170 were analyzed as UCG. Dropout rates were 10% for SMT and SMT+BT at week 3, 31% and 30% at week 4, and 49% and 50% at week 12. Adherence rates above 80% were reached in both interventions; 134 adverse events occurred. Intervention effects compared to UCG were found for pain intensity (SMT, P=.011, effect size d=0.41), disability (SMT+BT, P=.020, d=0.41), and peak torque (SMT, P=.045, d=0.38; SMT+BT, P=.019, d=0.44), with overall small effect sizes. Conclusions: Participants were highly adherent to the sensorimotor exercise, but showed increased dropout rates, particularly during home-based training. Both interventions proved to be feasible, and although only SMT showed an increased effect on pain intensity compared to UCG, the SMT+BT showed positive effects on disability. Both interventions led to increases in strength, indicative of a neuromuscular adaptation.