The Neugebauer approach to modeling color cmy halftones generally has to be modified to correct for the Yule—Nielsen light scattering effect. The most common modification involves the Yule—Nielsen n factor. A less common, but more fundamentally correct modification of the Neugebauer model involves a convolution of the halftone geometry with the point spread function, PSF, of the paper. The probability model described in the current report is less complex than the PSF convolution approach but is still much less empirical than the Yule—Nielsen n model. The probability model assumes the Neugebauer equations are correct and that the Yule—Nielsen effect manifests itself in a variation in the XYZ tristimulus values of the eight Neugebauer primary colors as a function of the amounts of c, m, and y printed. The model describes these color shifts as a function of physical parameters of the ink and paper that can be measured independently. The model is based on the assumption that scattering and absorption probabilities are independent, that the inks obey Beer–Lambert optics, and that ink dots are printed randomly with perfect hold-out. Experimentally, the model is most easily tested by measuring the shift in the color of the paper between the halftone dots, and experimental microcolorimetry is presented to verify the model.