This paper looks into the differences and uncertainties in determining the impact of “metals” emissions
on human health, in Life Cycle Impact Assessment (LCIA). Metals are diverse substances, with
different properties and characteristics, considered important in LCIA because of their toxicity to
humans and ecosystems. First, we defined a list of the most significant metals in terms of impacts on
human health. This was done according to precise criteria accounting for both physical and toxic
properties of the metals. Second, we performed a LCIA on different key processes using various
existing LCIA methodologies and including also USEtox: the recently developed consensus-model for
LCIA. Last, we compared the results in relative terms using a contribution analysis.. The analysis
showed poor or no agreement between the methods: the relative contribution of each metal and of the
metals in total to the total impact on human health changes greatly according to the LCIA method used.
These differences are due mainly to the number of metals included in each method and to the technique
used to calculate the characterization factors. Results obtained with USEtox show no apparent
correlation with results calculated with other methods. At present time USEtox is recommended as the
best model for LCIA on human toxicity, but mainly because of the large consensus behind it, because
its uncertainties regarding metals are still high. The study gives a good and simple overview regarding
the way different methods address the impact assessment for metals, and helps in the interpretation of
LCIA results for actual LCA studies where metal emissions are involved.