Bioeconomy development is increasingly promoted as a way to mitigate the impacts of fossil fuel consumption on climate change. However, modern intensive agriculture remains heavily dependent on industrial inputs, compromising its energy efficiency in providing biomass to produce the materials and energy society needs. Agroecology aims to reduce input use by enhancing ecosystem services. While ecosystem service assessment is an active field of research, existing approaches primarily focus on the economic and biophysical valuation. Yet, their potential to improve energy efficiency and autonomy remains largely overlooked. We built on existing energy assessment frameworks for agroecosystems to explicitly integrate ecosystem services and natural capital. Using a multi-indicator approach, we developed a set of indicators that capture energy consumption, efficiency, ecosystem service contribution, yield performance, and natural capital dynamics in agroecosystems. We tested this approach on 34 different arable agroecosystems north-eastern France. Our results show that organic agroecosystems yielded the best performance in terms of energy efficiency and agroecological integration, but suffer from low outputs due to lower yields than conventional agroecosystems. In the latter, fertilisers and irrigation were the main contributors to the energy footprint, but they enabled the highest absolute outputs. This framework provides insights into key agronomic levers for improving efficiency and autonomy, such as integrating leguminous leys in crop rotations.