The popular colorimetric assay for urinary 5-hydroxyindole-3-acetic acid based on the reaction with 1-nitroso-2-naphthol has been criticized for lack of specificity. Goldenberg (Clin. Chem. 19: 38-44, 1973) proposed a modification of this method, in which 2-mercaptoethanol is used to discharge unwanted chromophores. We compared results by both the unmodified and modified colorimetric methods with those by a "high-performance" liquid-chromatographic method, with fluorometric detection. Agreement among all three assays was excellent in the range of 50 to 600 mg/24 h, but there were considerable differences in the range 0 to 30 mg/24 h between results by the unmodified colorimetric assay and the other two. In this range, results by the modified assay (y) correlated well with those by the chromatographic method (x), although the values were slightly lower (y = 0.85x + 0.11, r = 0.9303). Using the modified colorimetric assay to analyze 40 normal specimens, we defined a normal range of 0 to 6 mg/24 h). We recommend the modified colorimetric assay for use in the high-volume laboratory and either this assay or the liquid-chromatographic method for the smaller laboratory.