The integration of digital twins (DTs) and dynamic life cycle assessment (LCA) offers a transformative approach to enhancing environmental sustainability in manufacturing systems. This paper proposes a conceptual framework to guide the development and implementation of DT-enabled dynamic LCA, emphasising system embeddedness – the integration of physical systems, virtual models, data, and performance objectives across multiple scales of manufacturing systems. This holistic approach enables the identification of environmental hotspots dynamically and the real-time monitoring of targeted optimisation strategies. The framework comprises three core pillars: (1) a multiscale DT architecture for real-time data collection and representation from diverse sources; (2) a dynamic LCA by leveraging real-time data feed from DTs and integrated with simulation modules in the DT system; and (3) a continuous feedback loop between DT’s simulation and optimisation with LCA for predicative decision support and operational adjustments in the physical manufacturing system. This study contributes a roadmap for leveraging DT technology and dynamic LCA to drive real-time environmental improvements across the manufacturing system, empowering manufacturers to make real-time data-driven decisions that promote a more sustainable and resource-efficient system.