Background: Transcutaneous electrical nerve stimulation (TENS) has been a commonly used modality to relieve aches and pain for over 40 years. Commercially available devices provide multiple therapy modes involving a different combination of frequency and pulse width with intensity. While frequency sets sensation, intensity helps determine tolerability, longer pulse width is reported to induce a feeling of deeper stimulation. In fact, longer pulse width has been empirically shown to deliver current into deeper tissues, but in context of other electrical stimulation modalities. The goal of this study was to unpack the relationship between pulse width and activation depth in TENS. Methods: A highly realistic, anatomically-based, 3D finite element model of the forearm was used to simulate the electric field (E-field) distribution, as the pulse width is varied. A typical titration-guided mechanism was used to obtain the strength-duration (S-D) curves of a sensory McIntyre-Richardson-Grill (MRG) axonal model simulating the pain-transmitting A-delta fibers. The pulse widths tested ranged from 30 μs to 495 μs. Results: As expected, shorter pulse widths required more current to achieve activation, resulting in a larger E-field. The S-D curve of the target median nerve indicates a rheobase of 1.75 mA and a chronaxie of 232 µsec. When the applied currents are the same, shorter pulse widths result in a smaller volume of tissue activated (VTA) compared to the longer pulse widths. A 21 fold difference in VTA was found between the longest and shortest pulse widths considered. We observed a linear relationship between pulse width and activation depth for the conditions tested in the study. Conclusion: Our findings highlight the impact of pulse width on activation depth. While choice of a given therapy mode is usually based on an ad-hoc desirable sensation basis, medical professionals may consider advocating a certain therapy mode based on the depth of the intended target nerve.