This manuscript addresses the consequences of the common practice of assuming that the gravel fraction of sediments does not participate in sorption reactions and thus sorption quantified by the distribution coefficient (K(d)) construct can be estimated from laboratory tests on sediments less than 2 mm size fraction. However, this common assumption can lead to inaccurate estimates of the mobility and sorption affinity of many radionuclides (e.g., Tc, U, and Np) on gravel dominated sediments at the Hanford Site and other locations. Laboratory batch sorption experiments showed that the distribution coefficients measured using only sediment less than 2 mm size fraction and correcting for inert gravel fraction were not in agreement with those obtained from the bulk sediments including gravel (larger than 2 mm size fraction), depending on the radionuclide. The least reactive radionuclide, Tc had K(d) values for bulk sediment with negligible deviations from the inert gravel corrected K(d) values measured on less than 2 mm size fraction. However, differences between measured K(d) values using sediment less than 2 mm size fraction and the K(d) values on the bulk sediment were significant for intermediately and strongly reactive radionuclides such as U and Np, especially on the sediment with gravel fractions that contained highly reactive sites. Highly reactive sites in the gravel fraction were attributed to the presence of Fe oxide coatings and/or reactive fracture faces on the gravel surfaces. Gravel correction factors that use the sum of the K(d)(,<2 mm) and K(d)(,>2 mm) values to estimate the K(d) for the bulk sediment were found to best describe K(d) values for radionuclides on the bulk sediment. Gravel correction factors should not be neglected to predict precisely the sorption capacity of the bulk sediments that contain more than 30% gravel. In addition, more detailed characterization of gravel surfaces should be conducted to identify whether higher reactive sorbents are present in the gravels.