Warning of workers in dangerous situations is crucial. With the aim of deriving practical parameters for an electrocutaneous warning stimulation, we explore the design dimensions of pulse intervals, amplitudes, and locations of electrocutaneous stimulation in a study on healthy volunteers. Using biphasic rectangular current pulses on the upper right arm of 81 healthy participants, they evaluated temporal perception with varying intervals, ranging from 200 ms down to 0.5 ms, categorizing it as ’Individual pulses’, ’Pulsating’, ’Vibrating’, or ’Continuous’. Next, we tested nine amplitude levels. Participants rated the perceived amplitude on a scale from 1 to 9 after a training phase. Finally, we presented five consecutive stimulation pulses in a pseudo-random order at eight electrode pair positions, asking participants to report the stimulated electrode pair. Participants perceived electrocutaneous pulses as ’Individual pulses’ for median intervals above 74 ms, as ’Pulsating’ between 44 ms and 74 ms, as ’Vibrating’ between 12 ms and 44 ms, and as ’Continuous’ below 12 ms. Pulse intervals below about 1 ms were perceived as weak and at about 5 ms as inconvenient, rendering these intervals less suitable for the design of a warning pattern. The median reported amplitudes [25%-75%-percentile] for presented amplitudes 1 to 9 are: 1[1–1], 2[2–3], 3[2–4], 3[3–4], 4[3–5], 5[4–6], 6[4–7], 7[5–8] and 7.5 [6–8] indicating a linear relationship between presented and perceived amplitude. These results suggest that the stimulation amplitude may be incorporated into a structured stimulation pattern. The majority of the electrode pair locations were reported correctly (64.3%–86.6%) or within the two neighboring electrode pairs (98%–99.7%). We conclude that the determined pulse intervals combined with the differentiability of locations offer the basis for designing a warning signal. Our research lays the groundwork for developing suitable signals for wearable electric warning devices.