AbstractSource-sink methods of calculation are compared with a set of experiments with mixed-fuel lattices in a heavy-water moderator. The theoretical model used was C. N. Klahr's modification of the Feinberg-Galanin theory. A comparison of calculation and experiment was made both for thermal-neutron density ratios within the lattice components, and for critical geometric bucklings (radial and vertical separately). Agreement was good in the case of thermal-neutron density ratios and rather good for radial bucklings. The moderate disagreements between calculated and experimental critical vertical bucklings were ascribed primarily to errors in the cell codes used to generate input for the mixed lattice calculations.