We compare the Roche ion-exchange column chromatography and the Merck antibody inhibition reaction for CK-MB fractionation in 51 sera. Measurements of total CK and CK-MB activities must be done under the same conditions for each method in order to correlate the results. 2. A decisional value must be used for the interpretation of CK-MB results. We have used 10% for the inhibition assay and 3% for the chromatography procedure. The use of a percentage should be preferred to use of CK-MB activity alone. 3. When the % of CK-MB was established for the 51 patients only 4 results disagreed between the two methods. Three of these could be explained by a lack of sensitivity of the column chromatography procedure. 4. The antibody assay produces reliable results. Since CK-BB and CK-MB are simultaneously measured, the method is therefore prone to interference by CK-BB when present in serum. The assay is greatly affected by the presence of adenylate kinase in serum. It is not necessary to run a serum blank with this procedure when the serum in pre-incubated for 7 minutes with the reagents. 5. The Roche method also produces reliable results but offers less sensitivity when total CK remains in the normal range. The procedure is much less affected by the presence of adenylate kinase.