The objective of this study is to develop an experimental protocol to define the range of modulation for different sensory modalities elicited by transcutaneous electrical nerve stimulation (TENS) in amputees with evoked tactile sensation (ETS). Modulation ranges of sensory modalities, such as vibration, buzz, tingling etc., are essential for designing a coding strategy for incoming sensory information from prosthetic hands for each amputee. Here, the modulation ranges of different modalities in pulse width at fixed stimulus frequencies were evaluated. Four healthy subjects and three transradial amputee subjects were recruited to participate in this preliminary test. A single skin site in healthy subjects and five finger areas of the projected finger map (PFM) on the stump skin in amputee subjects were stimulated for evaluation. Then, a finger identification test was conducted to show the feasibility of this sensory coding strategy in one amputee subject. The sensitivity of different sensory modalities was also measured to quantify the minimum pulse width change that subject could distinguish. Results showed that vibration and buzz sensations had wider modulation ranges in pulse width for both healthy and amputee subjects at 20Hz and 50Hz of stimulation, respectively. The average accuracy of finger identification was 91.66% in the amputee subject. The minimum pulse width changes distinguishable for both vibration and buzz sensation were below 20 (us). Results indicated that the protocol of evaluating pulse width modulation ranges for different sensory modalities was effective, and the coding strategy could provide accurate finger-specific sensory information for amputees with ETS.Clinical Relevance— This protocol establishes a guideline for customizing the coding strategy of evoked tactile sensory feedback for each amputee.