Abstract: For over 3 million years hominins held stone-cutting tools in the hand, gripping the portion of tool displaying a sharp cutting edge directly. During the late Middle Pleistocene human populations started to produce hafted composite knives, where the stone element displaying a sharp cutting edge was secured in a handle. Prevailing archaeological literature suggests that handles convey benefits to tool users by increasing cutting performance and reducing musculoskeletal stresses, yet to date these hypotheses remain largely untested. Here, we compare the cutting performance of hafted knives, ‘basic’ flake tools, and large bifacial tools during two standardized cutting tasks. Going further, we examine the comparative ergonomics of each tool type through electromyographic (EMG) analysis of nine upper limb muscles. Results suggest that knives (1) recruit muscles responsible for digit flexion (i.e. gripping) and in-hand manipulation relatively less than alternative stone tool types and (2) may convey functional performance benefits relative to unhafted stone tool alternatives when considered as a generalised cutting tool. Furthermore, our data indicate that knives facilitate greater muscle activity in the upper arm and forearm, potentially resulting in the application of greater cutting forces during tool use. Compared to unhafted prehistoric alternatives, hafted stone knives demonstrate increased ergonomic properties and some functional performance benefits. These factors would likely have contributed to the invention and widespread adoption of hafted stone knives during the late Middle Pleistocene.