Structural health monitoring strategies in civil engineering increasingly leverage digital twins, supporting predictive maintenance and data-driven decision-making. Despite the widespread adoption of digital twin applications in structural health monitoring (SHM), there is a lack of agreement on a common definition of digital twins. Although a significant number of digital twin definitions have been proposed and a plethora of reviews have been published, little emphasis has been given to digital twin architectures and internal elements. This paper presents a multivocal review of digital twins in civil engineering, aiming to provide a panorama of the digital twin landscape in civil engineering with explicit insights into the architectures and internal elements used in digital twin applications. From a methodological standpoint, the review follows a twofold approach that encompasses (i) peer-reviewed, indexed literature (white literature) as well as (ii) non-indexed sources (gray literature) that include industrial digital twin applications. Furthermore, drawing from the review results, a digital twin definition is formulated and a generic digital twin reference architecture is proposed. It is expected that the definition and the generic reference architecture proposed in this study may serve as a blueprint for digital-twin-based SHM applications, with significant implications for researchers, practitioners, and policymakers in structural health monitoring. © 2024, NDT. net GmbH and Co. KG. All rights reserved.