Albedos were experimentally determined for a spectrum of fast fission neutrons incident on a 24- × 24- × 9-in. concrete slab and a 24- × 24- × 6-in. cold-rolled steel slab. The experimental dose albedos were compared with those from Oak Ridge National Laboratory (ORNL) 05R Monte Carlo calculations and with orders-of-scattering calculations performed using readily available multigroup cross sections. The source of fast neutrons for the experimental studies was the Kansas State University Triga Mark II reactor; the direct and reflected neutron spectra were detected using a 2- × 2-in. liquid-scintillation spectrometer system. A computer code has been developed to normalize the experimental reflected spectra to a reference direct neutron beam measurement, to calculate the experimental dose albedos, and to evaluate the Monte Carlo dose albedos in terms of the experimental conditions. A comparison of the Monte Carlo and experimental results showed similar trends and structure, but the experimental dose albedos were smaller than the calculated ones by an average of 70 to 80%. The large energy bin widths and inaccuracies in the cross sections (errors up to 65%) used in the Monte Carlo calculations have contributed to the differences between the calculated and experimental results. Orders-of-scattering calculations for steel were performed using the ORNL DLC-2B 99-group cross-section set and similar calculations were performed for concrete using the DLC-9/FARS group cross-section set. In general, the orders-of-scattering calculated reflected spectra compare well with the experimental fast-neutron reflection spectra. Discrepancies arise in the orders-of-scattering calculated results at particular energies and scattering angles due to deficiencies identified in the multigroup compilations.