The speciation of aluminium in serum and in water is discussed. Ultramicrofiltration and HPLC techniques have been used to show that > 90% of the aluminium in serum is non-ultrafiltrable and bound to the protein transferrin. Regarding the toxicity and speciation of aluminium in natural waters the literature is plagued by disparities. The present lack of designed protocols and reference waters for the determination of the toxic aluminium species is thought to play a key role in producing these disagreements.