The oral cavity is richly innervated with mechanoreceptors that provide information about bolus characteristics. It is hypothesized that this information is important for modulating tongue and jaw movement to achieve coordinated, efficient chewing and swallowing. To test this hypothesis, we performed an intraoral lidocaine nerve block on multiple sensory branches of the trigeminal nerve to temporarily silence sensory signals from mucosal and periodontal mechanoreceptors while preserving motor and proprioceptive signals to and from the jaw and tongue muscles. We collected biplanar videoradiographic data before and immediately after the nerve block and quantified jaw and tongue kinematics using XROMM (n=2 Macaca mulatta, both male). We found that in the absence of tactile sensation average feeding sequence duration, in terms of number of cycles, increased. Chewing after the nerve block was more rhythmic (i.e., more consistent cycle durations) and the temporal relationship between gape and tongue tip anteroposterior reversal became more stereotyped. Furthermore, average gape magnitude remained consistent but within-sequence variation in gape magnitude decreased significantly after the block. On the whole, our results suggest that proprioceptive inputs from jaw and tongue muscle spindles are sufficient for successful feeding but that tactile and periodontal inputs play a key role in modulating kinematics for efficient food breakdown, as evidenced by longer feeding sequence durations as well as reduced variation in cycle duration and gape magnitude in the nerve block condition.