[Purpose] This study compared the effectiveness of stabilization and McKenzie exercises on pain, disability, and thickness of the transverse abdominis and multifidus muscles in patients with nonspecific chronic low back pain. [Subjects] Thirty patients were randomly assigned into two groups: the McKenzie and stabilization exercise groups. [Methods] Before and after intervention, pain, disability, and thickness of the transverse abdominis and multifidus muscles were evaluated by visual analogue scale, functional rating index, and sonography, respectively. The training program was 18 scheduled sessions of individual training for both groups. [Results] After interventions, the pain score decreased in both groups. The disability score decreased only in the stabilization group. The thickness of the left multifidus was significantly increased during resting and contracting states in the stabilization group. The thickness of the right transverse abdominis during the abdominal draw-in maneuver, and thickness of the left transverse abdominis during the active straight leg raising maneuver were significantly increased in the stabilization group. The intensity of pain, disability score, thickness of the right transverse abdominis during the abdominal draw-in manouver, and thickness of the left transverse abdominis during active straight leg raising in the stabilization group were greater than those on the Mackenzie. [Conclusion] Stabilization exercises are more effective than McKenzie exercises in improving the intensity of pain and function score and in increasing the thickness of the transverse abdominis muscle.