The widely used thiobarbituric acid technique for the quantitation of N-acetylneuraminic acid (NANA) was improved to eliminate the interference of the ubiquitous 2-deoxyribose. The 2-deoxyribose chromogen was completely removed by cyclohexanone extraction at pH 5.6-6.0. After readjusting the pH to 1.7-1.9, the chromogen representative of NANA was quantitatively extracted with cyclohexanone. All other aspects of the original technique (L. Warren 1959 J. Biol. Chem. 534, 1971-1975) remained unchanged. The technique has been applied to determine total as well as neuraminidase-susceptible NANA in the preparation of the immunogen (neuraminidase-treated myeloblasts) utilized to stimulate specific immunity in patients with myeloblastic leukemia and certain solid tumors; NANA levels significantly affect immunogenicity. Data obtained from a variety of tumors using pH-dependent extraction as compared to the thiobarbituric acid method, isoamyl alcohol extraction, and ion-exchange purification showed that 2-deoxyribose interference may cause as much a two- to threefold error in the quantitation of NANA.