Online condition monitoring is an area of active research that may enable optimized scheduling, maintenance, and safety of nuclear power plant components, reducing unnecessary derates while simultaneously improving operational capacity. Digital twins (DTs) are one avenue to conduct online condition monitoring and are currently being explored by national laboratories and universities alike. DTs for online condition monitoring are, in essence, state concurrent models that emulate a physical process that predicts a parameter and compares it against a measured value. A DT’s goal is to provide additional insights by combining and interpreting various sources of information for preventative maintenance scheduling optimization or early fault detection. Condition monitoring for DTs are projected to be valuable for meeting requirements under 10 CFR 50.55a, “Codes and Standards,” and 10 CFR 50.65, “Requirements for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants.” However, DT technologies are still under significant development, and the process for developing a DT for condition monitoring have not been formalized. Therefore, in this work, we present an initial framework for developing a DT, discuss and review the various challenges and considerations for DT deployment, and identify the opportunities that a DT can improve. The presented framework is intended to help developers formulate a strategy when approaching DT development for condition monitoring. A DT use case for a reactor coolant pump is presented to demonstrate the proposed framework.