The interaction of haemin with human serum albumin has been reexamined. The absorption spectrum of the bound haem is identical with that of uncomplexed monomeric haemin in solution, and it is suggested, on the basis of an interaction of albumin with iron-free protoporphyrin IX, that the iron is not implicated in the interaction with the protein. A ferric cyanide derivative, and a ferrous haem derivative of methaemalbumin can be recognised, but not azide or fluoride derivatives. The bound haemin gives rise to extrinsic Cotton effects, which are different in detail in the derivatives, and in the complex with protoporphyrin IX. Spectrophotometric titrations show that there is one strong binding site for haemin and several weaker sites. The latter are associated with optical activity opposite in sign to that of the strong complex. The binding profiles are little affected by p H over a wide range, by ionic strength or by the presence of 40°/, (vlv) dimethylsulphoxide/water, in which the free haemin is monomeric. The binding of haemin to albumin has been followed by spectrophotometry, circular dichroism and fluorescence. The binding of haemin quenches the protein fluorescence, which progressively changes in character from tryptophan to tyrosine type. The haem a t the primary binding site thus strongly quenches the tryptophan specifically. From fluorescence titrations a t a range of protein concentrations, binding isotherms were constructed, and gave an association constant for the strong site of 50 pM-I. From binding isotherms based on absorption measurements we can infer the existence of a t least four sites with much lower binding constants. A long-chain fatty acid anion was found to compete with haemin only for the weaker binding sites, so that binding beyond one mole per mole of protein can be essentially eliminated. The open-chain tetrapyrrole, bilirubin, was found, in agreement with earlier work, not to compete with haemin, as regards the strongest binding sites of either ligand. Between the weaker sites, however, evidence of competition was obtained.Haemin will bind to human serum albumin under physiological conditions to give a spectroscopically identifiable complex known as methaemalbumin [l]. This same complex is also formed when haemoglobin is incubated with albumin [2], or in vivo in acute haemolytic conditions, such as blackwater fever [3], haem being in both cases transferred from the globin. For this reason it appears in the blood after haemolysis only after the haptoglobin has been saturated [4]. The binding of haemin to albumin is species-specific, and does not appear to occur to the same extent, or perhaps at all, with many animal albumins [2]. The nature of the interaction with human serum albumin has not apparently received any attention since the work of Rosenfeld and Surgenor [i] in 1950, in which the existence of two binding sites of equal affinity was inferred. Developments in the intervening years in both the techniques and theory of proteinligand systems suggest that the system should be re-i...