Grain-size dependencies in fallout radionuclide activity have been attributed to either increase in specific surface area in finer grain sizes or differing mineralogical abundances in different grain sizes. Here, we consider a third possibility, that the concentration and composition of grain coatings, where fallout radionuclides reside, controls their activity in fluvial sediment. We evaluated these three possible explanations in two experiments: 1) we examined the effect of sediment grain size, lithology, mineralogy, and composition of the acidextractable materials on the distribution of 7Be, 10Be, 137Cs, and unsupported 210Pb in detrital sediment samples collected from rivers in China and the United States, and 2) we periodically monitored 7Be, 137Cs, and 210Pb retention in samples of known composition exposed to natural fallout in Ohio, USA for 294 days.Acid-extractable materials (made up predominately of Fe, Mn, Al, and Ca from secondary minerals and grain coatings produced during pedogenesis) are positively related to the abundance of fallout radionuclides in our sediment samples. Grain-size dependency of fallout radionuclide concentrations was significant in detrital sediment samples, but not in samples exposed to fallout under controlled conditions. Mineralogy had a large effect on 7Be and 210Pb retention in samples exposed to fallout, suggesting that sieving sediments to a single grain size or using specific surface area-based correction terms may not completely control for preferential distribution of these nuclides.We conclude that time-dependent geochemical, pedogenic, and sedimentary processes together result in the observed differences in nuclide distribution between different grain sizes and substrate compositions. These findings likely explain variability of measured nuclide activities in river networks that exceeds the variability introduced by analytical techniques as well as spatial and temporal differences in erosion rates and processes. In short, we suggest that presence and amount of pedogenic grain coatings is more important than either specific surface area or surface charge in setting the distribution of FRNs. Cs, and 210 Pb in river sediment (GCA manuscript GCA-D-16-0143R2).We are again appreciative of the through and thoughtful review from one reviewer who has previously reviewed the manuscript, particularly with respect to measurements of 7 Be in the "natural delivery" experiment. The reviewer has helped us to clarify the main points of that experiment and to better explain the sources of error in those data. We have made minor changes to the manuscript to reflect these concerns.Below we have enclosed a detailed response point by point to each reviewer concern. Reviewer comments are in black and our responses are in blue.The manuscript is now 7,239 words from the introduction through the acknowledgements and includes a 308 word abstract, 8 figures, 1 table, 66 references, and supporting information including text, 13 figures, and 11 tables. This manuscript is not under con...