Reproducibility and methodological robustness are of major concern in research today, also with respect to autonomic measures. Quantification of skin conductance responses (SCRs), for instance, relies on response characteristics such as response onset (i.e., latency), which were established more than four decades ago by using paper-pencil methods. Since then, data acquisition has advanced to digital methods, improving sampling rates up to 100,000 samples/second and thereby improving resolution and accuracy. Here, SCR latency as a major characteristic for defining an appropriate response window is revisited by using state-of-the-art equipment. Furthermore, SCR latencies are investigated across different stimulus modalities (tactile, auditory, visual) and with respect to their temporal dynamics over the course of a learning experiment (i.e., fear conditioning). The established response latency criteria were largely confirmed even though results did suggest an even narrower and potential stimulus modality-specific latency criterion. Exploratory analyses investigating individual differences in SCR latencies provided first evidence for a role of sex and cognitive effects (i.e., contingency awareness) while dispositional negativity as well as other personality traits did not affect SCR latencies. Consequently, SCR latencies might have an informative and discriminative value beyond a solely criterial function for defining response windows. The current findings may help to improve the rigor of using SCRs and suggest that SCR latency as a descriptive measure warrants further investigation.