Despite its well‐documented limitations, colorimetry has been commonly used for the d‐xylose test in the diagnosis of malabsorption syndrome (MAS). With a possibility of overcoming its limitations, the use of ¹H NMR spectroscopy for D‐xylose test is explored herein. Urine samples from 35 adults with suspected MAS were obtained before and after oral ingestion of D‐xylose. The diagnosis of MAS was based on fecal fat (72 h excretion using Van de Kamer's technique, normal < 7 g/24 h and/or Sudan III stain of spot stool specimen, normal≤10 droplets/high power field) and/or endoscopic duodenal biopsy. Urinary excretion of D‐xylose over 5 h after consumption of 5 g D‐xylose, using both colorimetry and NMR was compared (normal≥1 g/5 g/5 h). In vitro experiments on the standard specimens of D‐xylose were also performed independently using both methods. Colorimetry showed a lower value for the quantity of D‐xylose excreted in urine than NMR [median 0.73 (0.17–1.89 g) vs 1.37 (0.17–3.23 g), respectively; p<0.0001, Wilcoxon's signed ranks test]. Colorimetry and NMR correctly diagnosed 11/12 and 10/12 (p=N.S.) patients with MAS and 14/23 and 20/23 (p<0.05) without MAS, respectively. Sensitivity and specificity of colorimetry and NMR were 91.6 and 60.7% vs 83.3 and 86.9%, respectively. In in vitro experiments, the values obtained for standard xylose using NMR showed a maximum error of 7%, whereas the colorimetric method showed 20%. The NMR method is simple and may be more accurate for the D‐xylose absorption test. Colorimetry was found to be inferior as compared with NMR due to its low specificity. Copyright © 2004 John Wiley & Sons, Ltd.