Study Design Controlled laboratory study. Objectives To clarify whether differences in surface stability influence trunk muscle activity. Background Lumbar stabilization exercises on unstable surfaces are performed widely. One perceived advantage in performing stabilization exercises on unstable surfaces is the potential for increased muscular demand. However, there is little evidence in the literature to help establish whether this assumption is correct. Methods Nine healthy male subjects performed lumbar stabilization exercises. Pairs of intramuscular fine-wire or surface electrodes were used to record the electromyographic signal amplitude of the rectus abdominis, the external obliques, the transversus abdominis, the erector spinae, and lumbar multifidus. Five exercises were performed on the floor and on an unstable surface: elbow-toe, hand-knee, curl-up, side bridge, and back bridge. The EMG data were normalized as the percentage of the maximum voluntary contraction, and data between doing each exercise on the stable versus unstable surface were compared using a Wilcoxon signed-rank test. Results With the elbow-toe exercise, the activity level for all muscles was enhanced when performed on the unstable surface. When performing the hand-knee and side bridge exercises, activity level of the more global muscles was enhanced when performed on an unstable surface. Performing the curl-up exercise on an unstable surface, increased the activity of the external obliques but reduced transversus abdominis activation. Conclusion This study indicates that lumbar stabilization exercises on an unstable surface enhanced the activities of trunk muscles, except for the back bridge exercise. J Orthop Sports Phys Ther 2010:40(6):369–375.doi:10.2519/jospt.2010.3211