In this article, a spectrophotometric color maching algorithm based on Stearns-Noechel model is proposed. This algorithm was run to predict recipes for 48 viscose blends. Color differences between the original blend samples and the calculated were expressed in CIELAB units (10 degrees standard observer). M (the empirical constant in Stearns-Noechel model) value was determined by median analysis. When M equals to 0.09, the best fit was obtained for three-components fiber blends. In that case, the maximum color difference is 1.22 CIELAB units and the average computed color difference is 0.56 CIELAB units for 36 three-components fiber blends under D65 illuminant. When M is from 0.03 to 0.06, the best fit was obtained for four-components fiber blends. In that case, the maximum color difference is 4.48 CIELAB units and the average computed color difference is 1.02 CIELAB units for 12 four-components fiber blends under D65 illuminant. It is demonstrated that the algorithm can be used in color matching of fiber blends. (c) 2009 Wiley Periodicals. Inc. Col Res Appl, 34, 108-114, 2009; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/col.20478