<br/>Life Cycle Assessment quantifies the environmental impacts from emissions and resources consumption of human activities. Uncertainty in modelling natural processes and ecological regulation challenges the prediction of effects from further pressures. Ecosystems’ health and adaptation capacity may have also been altered by past impacts. Model frameworks are usually built on stability, linearity of causality and expectation of a safe return to stable states if the stressor is minimised. However, the command-and-control paradigm has resulted in the erosion of natural resources and species diversity. Ecosystem-related impacts are traditionally benchmarked by potential loss of biological diversity as Potentially Disappeared Fraction of species (PDF) integrated over area and time, building on the biological sensitivity of species in each receiving ecosystem. For consistency among Life Cycle Impact Assessment (LCIA) methods midpoint indicators are shown in Potentially Affected Fraction of species (PAF), which implicitly suggests reversibility to previous stable states. Currently applied conversion factors from midpoint to endpoint (species loss, as PDF) range from 10 (NOEC-based), 2 (chronic EC50-based) or 1 (assuming that continuous stress affects reproduction rate), but these are all based on biological/physiological responses and do not add a true ecological component to the impact. Such factor simply changes the HC50 by 1 or 0.3 log units. A stressor with equal intensity in two differently disturbed ecosystems (close or distant to a threshold) and sharing similar biological communities should not result in, necessarily, the same impact potential. We suggest the introduction of an ecological term in the Effect Factor of the characterisation modelling for ecosystem quality-related indicators in LCIA. An application to a marine eutrophication indicator will be presented to show how impacts from nitrogen emissions vary with the individual receiving ecosystems’ health, by defining proxies for ecosystem’s state and resilience. These, express the pressure on the system and its propensity for regime shifting. Ultimately, the ecosystem’s capacity to tolerate the pressure, to adapt to the stress and minimise its effects should complement the biological response. In our view, adding an ecosystem-based approach to the damage estimation can positively contribute to the environmental relevance and spatial differentiation of the results.