Parameter measurements and calculations in two D2O-moderated thorium-uranium critical assemblies are described in detail. The first, designated ETA-I, contained 6.7 wt% 235UO2-ThO2 fuel rods 0.66 cm in diameter, clad in aluminum. The second assembly, ETA-II, contained 3.0 wt% 233UO2-ThO2 fuel rods 1.09 cm in diameter, clad in Zircaloy-2. A relatively hard spectrum was obtained in both lattices. Because leakage was small and the central flux spectra were closely asymptotic, these assemblies provided a cleanly analyzable test of important reaction cross sections. Parameters measured in ETA-I were 232Th fission/235U fission (δ02), 232Th capture/235U fission (CR*), and epithermal/thermal-neutron ratios for 235U fission (δ25) and 232Th capture (p02). Corresponding measurements with 233U were made in ETA-II; δ02 was measured directly in fuel and ThO2 pellets by observing specific fission product gamma rays with a high resolution Ge(Li) detector. With this system, 232Th capture gamma-ray activity was also measured directly in a fuel pellet. The epithermal/thermal-neutron ratios were measured by the thermal-neutron subtraction technique relative to 164Dy, and CR* was measured relative to its thermal-neutron value in a very soft spectrum. A three-dimensional adjoint Monte Carlo program was used to calculate foil flux perturbations and cadmium cutoff energies. Supplemental fast/epithermal-neutron spectrum comparisons were made between the ETA assemblies and the 4/1 TRX critical assembly by means of the following activation detectors: 197Au(n,γ), 55Mn(n,γ), 235U(n,f), 238U(n,f), 27Al(n,α), and 96Zr(n, γ). The lattices were analyzed explicitly with a full energy range Monte Carlo program having a detailed cross-section description. ENDF/B Version 2 and light-water breeder reactor (LWBR) cross-section sets were used. On the whole, agreement with experiment was very good.