Protein engineering allows us to change the amino acid sequence of a protein in specific predetermined ways. The reasons for doing so include understanding the relationship between a protein's structure and its function; labelling the protein for structural or radiotracer experiments; or deliberate modification of a protein's functional properties, for industrial or clinical use. Formerly, protein engineers operated at the level of the protein, using a wide variety of chemical means to change its structure. Of these the most powerful is protein semisynthesis (14,15). Now, site-directed mutagenesis of cloned genes can change the codon for a specific amino acid, so that subsequent expression of the altered gene yields an engineered protein. This technique has become the method of choice for most proteins, except where a semisynthetic route is already well established or where the experimental goal is difficult to attain by genetic means, for example, the insertion of a non-coded amino acid (15). Our work has been concerned both with the development of techniques of protein engineering by semisynthesis (14), and their use in answering questions about structure-function relationships in the respiratory chain component cytochrome c (8). We continue projects designed to understand the mechanism of electron transfer between the protein and its physiological partners (1), the way in which the nature of the protein shell establishes the internal dielectric (10) and modulates the properties of the haem centre (2,4,5). In investigations of the way the protein interacts with phospholipid membranes we have revealed a novel type of interaction, the extended lipid anchorage(7) and have recently proposed a structural mechanism(3). We continue investigations of this phenomenon, not least because of the potential relevance to the process of apotosis.