The upcoming proliferation of Automated Vehicles (AVs) above Level 3 fundamentally disrupts established communicative conventions between road users and vehicles, and external Human–Machine Interface (eHMI) emerges as the primary channel for conveying vehicle intent to pedestrians, cyclists, and other drivers. Despite the standardisation of the marker lamp of AVs and the growing research activity since 2016, no internationally harmonized eHMI standard has been established as of early 2026. This Systematic Literature Review (SLR), following the PRISMA 2020 protocol, synthesizes empirical and technical findings from 149 peer-reviewed English journal articles spanning the decade from 2016 to 2026, from Web of Science, Scopus, IEEE Xplore, and Google Scholar. The review covers scenario design, test methodologies, participant demographics, eHMI modalities, and performance evaluation measures by mapping, coding, and presenting in a structured framework the reviewed papers. The key findings are as follows: (1) 13 use cases (50%, out of 26) still need to be further studied; (2) vehicle kinematics remain a primary determinant of crossing decisions, with eHMI signals serving as confirmatory rather than primary cues in many scenarios; and (3) cross-cultural and demographic diversity in empirical studies remains critically insufficient for global regulatory purposes. We synthesize these findings into a structured framework of evidence-based recommendations for international eHMI standardisation organizations and identify critical research gaps requiring resolution before comprehensive standards can be enacted.