The primary amino acid sequences of the 20 complement components and control proteins, found in plasma, and of many of the cell-surface molecules associated with the control of the complement system are known from recent cDNA cloning studies. This has indicated that most of these proteins contain a number of well-defined domains of 40-80 amino acids: between two and 30 domains of one type are found in some of the proteins, while others show a mosaic structure composed of more than three different types of domain. The same types of domain are found in a growing number of non-complement proteins such as blood clotting factors and certain cell adhesion molecules. The main purpose of a recent meeting was to assess and correlate the data emerging from structure comparison and prediction techniques and from functional and physicochemical studies of isolated domains and whole proteins.