Abstract Incorporating palaeontological data into the methods and formats already familiar to conservation practitioners may facilitate greater use of palaeontological data in conservation practice. Benthic indices (e.g. Multivariate-AZTI Marine Biotic Index; M-AMBI) already incorporate reference conditions and are a good candidate for integration. In simulations of living communities under constant and changing environmental conditions, we evaluate the capacity of death assemblage reference conditions to replicate M-AMBI values when used in place of reference conditions from the final ten generations of the simulation or all five hundred simulated generations. Reference conditions from all death assemblage scenarios successfully replicated correct remediation decisions in the majority of simulation runs with environmental change and stability. Variations in M-AMBI values were due to overestimated species richness and Shannon entropy values in the death assemblages but effects of changes to these parameters varied across scenarios, emphasizing the importance of evaluating multiple metrics. Time averaging was largely beneficial, particularly when environmental change occurred and short-term ecological observations (ten generations) produced incorrect remediation decisions. When the duration of time averaging is known, death assemblages can provide valuable longer-term perspectives with the potential to outperform temporally constrained baseline information from monitoring the living community.