Background Reduced lumbar multifidus (LM) muscle contraction has been observed in patients with low back pain (LBP). Clinicians often use various strategies to ensure LM activation, including tactile feedback and verbal instruction. However, the effects of tactile feedback on muscle activation have not been studied previously. Therefore, the purpose of this study was to investigate whether or not tactile feedback would increase LM muscle activity in adults with and without LBP. Methods Twenty asymptomatic adults and 20 patients with existing LBP completed the study. Two electromyographic (EMG) electrodes were applied to both sides of the LM at the L5 segment. EMG activity was collected three times at rest with and without tactile feedback, then five times during contralateral arm lifts with and without tactile feedback. The tactile feedback was applied by direct and continuous hand contact to the bilateral LM over the lumbosacral area. Lastly, two 5-second trials of maximum voluntary isometric contraction (MVIC) during a bilateral arm lift were performed. EMG activity collected at rest and during contralateral arm lifts was normalized to that collected during MVIC. Normalized EMG values of the right side of the asymptomatic group and the painful side of the LBP group were used for data analysis. Results Statistical analysis showed significantly decreased LM EMG activity with tactile feedback both at rest and during contralateral arm lifts compared to LM EMG activity without tactile feedback. There was no difference in LM EMG between the asymptomatic and the LBP groups. Conclusions The results of the study showed that adding tactile stimulation to verbal instruction appeared to provide an inhibitory effect on LM activity in both asymptomatic healthy adults and patients with LBP. Contrary to common belief, tactical feedback via direct hand contact may reduce LM muscle recruitment, and may lessen the desired treatment effect.