Background: Digital twins (DT) are the coupling of a real-world physical asset to a virtual representation to provide insight and actionable knowledge. The benefits of DT are considered to include improvements in reproducibility, reliability of interventions, increased productivity, as well as increased time for innovation. For instance, a DT could be used to boost agricultural productivity whilst also meeting various targets (e.g., biodiversity, sustainability). Or a DT could be used to monitor a cell culture, predict interactions, and make subtle adjustments to maintain the environment allowing researchers to conduct other work. Yet in developing DT two fundamental questions emerge: ‘What will the DT capabilities be?’ (i.e., the range of features and possible actions) and ‘What will the DT do?’ (i.e., which capabilities will it utilise). Methods: Here we discuss a theoretical framework for DTs developed during Wageningen University & Research’s Investment Programme on DTs that aims to answer these questions. Focusing on the Life and Environmental Sciences to help developers and stakeholders to agree on the capabilities, purpose, and goal of a DT. As well as identifying iterative design stages that may help set interim development goals such as a minimum viable product. Results: This framework defines a DT as sitting at one of five maturity, or capability, levels associated with specific types of DT: a status, an informative, a predictive, an optimisation, and an autonomous twin. Conclusions: The aim of DTs is to make better, data-driven, decisions yet there can be a gulf between expectations of what a Digital Twin will do and the reality. The five maturity levels outlined here can be used to first identify and communicate about the type of Digital Twin required for a particular project prior to DT development. Bridging the gap between what project leads, developers, and stakeholders envision the end-product will be.