Abstract:Precise tongue control is necessary for drinking, eating, and vocalizing1, 2. Yet because tongue movements are fast and difficult to resolve, neural control of lingual kinematics remains poorly understood. Here we combine kilohertz frame-rate imaging and a deep learning based neural network to resolve 3D tongue kinematics in mice drinking from a water spout. Successful licks required previously unobserved corrective submovements (CSMs) which, like online corrections during primate reaches3–10, occurred after the tongue missed unseen, distant, or displaced targets. Photoinhibition of anterolateral motor cortex (ALM) impaired online corrections, resulting in hypometric licks that missed the spout. ALM neural activity reflected upcoming, ongoing, and past CSMs, as well as errors in predicted spout contact. Though less than a tenth of a second in duration, a single mouse lick exhibits hallmarks of online motor control associated with a primate reach, including cortex-dependent corrections after misses.