We use an ion-exchange column-chromatographic technique for separating creatine kinase isoenzymes in serum, and occasionally observe what appears to be sustained increase in the MB fraction. Most patients whose sera show such behavior have myocardial disease, but not necessarily a recent myocardial infarction. Electrophoretic analysis of a small sampling of such sera revealed that the apparent MB migrates atypically, appearing distinctly between isoezymes MB and MM. In another electrophoretic system, the peak might easily be mistaken for MM. This unusual isoenzyme does not appear to be "macro" creatine kinase. In laboratories that use the ion-exchange technique, the possibility of a falsely positive MB value should be considered in subjects who show persistent increases together with normal or nearly normal values for total creatine kinase activity. A suitable electrophoretic method that clearly demonstrates this unusual isoenzyme should be used in such cases, for confirmation.