Experiments are described and reviewed demonstrating that the bound carbohydrates of glycoproteins of many forms of malignant cells differ from their normal counterpart. The difference involves many oligosaccharide groups and is essentially quantitative. The characteristics of the difference are discussed. Despite the consistency of the finding its significance is unknown because the function of bound carbohydrates is largely unknown. Some properties of protein-bound carbohydrates that may be of special relevance to malignancy and other pathological processes are considered. The array of structures found in the cell is highly complex but seems to be similar in man, hamster, mouse, chick and fish. On the other hand, the biosynthesis of these structures can be influenced and altered by the environment and by drugs; the cell is tolerant of variation in its bound carbohydrate; microheterogeneity of the carbohydrates is probably the rule rather than the exception; experiments to test the function of bound carbohydrate show only small effects. A role for the bound carbohydrates in evolution is proposed that is consistent with these characteristics. It is also postulated that altered, bound carbohydrates of most glycoprotein does not endanger the life of the cell but may be responsible for involvement and change of many processes some of which permit the malignant cell to divide persistently and to prosper.