With the development of techniques for the qualitative and quantitative study of isoenzymes (Latner and Skillen, 1961; Van der Helm, Zondag and Klein, 1963; Gerhardt, 1963 et al;Richterich, Schafroth and Aebi, 1963) some attention has been drawn to a study of isoenzymes in cancer. We have been studying isoenzymes of lactate dehydrogenase, leucine aminopeptidase, and alkaline phosphatase chiefly, using starch-gel electrophoresis. The gel buffer used is Tris-HCl 0·025 M, pH 8·6 and by this method five zones of activity of LDH and three zones for LAP can be well separated.A study of isoenzymes in cancer sera has, as have enzyme determinations, shown no ' cancer specific changes' though distributions are sometimes abnormal. For example, LAP shows three zones of activity, in contrast to the normal one, in cases of lymphosarcoma (Kowlessar, Haeffner and Riley, 1961), obstructive jaundice secondary to tumour, metastatic carcinoma to the liver and sometimes in cancer of the gastrointestinal tract.Other body fluids show abnormalities of distribution in cancer, for example, in malignant effusions LD 1 and LD 2 (the slowest migrating isoenzymes of LDH in our nomenclature) account for more than 30 % of the total LDH (Richterich, Locher, Zuppinger and Rossi, 1962). Characteristic distributions have also been reported in CSF LDH in some cases of metastatic carcinoma to the brain (van der Helm, Zondag and Klein, 1963).The isoenzymes of tumour tissues show a marked uniformity of pattern when similar amounts of activity are used. With regard to LDH, tumour tissues show predominantly LDl> LD 2 and LD a • When compared with normal adult tissues the tumour tissues show a shift to the slower migrating zones (those associated with aerobic glycolysis). Foetal tissue extracts are very similar to tumour tissues in isoenzyme distribution.Abnormal isoenzymes are sometimes detected in tumours though they are not common, and similar abnormalities can be detected in foetal tissues. Abnormal LDH isoenzymes are usually present in the LD a region and preliminary studies on crude isoenzyme preparations (Starkweather and Schoch, 1962) have shown a difference in Michaelis constants for tumour LD a and normal adult LD a which suggests that the former may be a structurally different protein.