To improve the quality of motor-task performance, the central nervous system (CNS) increases the precision of sensory information processing. Phase-dependent regulation of that precision during reach-to-grasp movements, reflected in the precision of transport-aperture coordination, decreases the cost of required neural computations. The precision of state estimation for the hand or tool is maximized by optimal integration of visual and proprioceptive information. In tool-mediated reaching, that integration results in biases regarding both the explicit and implicit perceptual estimates of hand position. Experimental data revealing such biases are analyzed to obtain valuable information about sensorimotor adaptation required for learning to use tools.