Fine particles may infiltrate through coarse alluvial beds and eventually saturate the subsurface pore space. It is essential to understand the conditions that lead to bed saturation, and to forecast the packing characteristics of saturated beds to assess the effect of excess fine sediment supply on a number of processes that occur in the stream–sediment boundary. To address this problem, in this study, a new method is introduced to predict the grain-size distribution for the saturated condition, and the resulting porosity decrease, given the characteristics of the bed and the supplied sediments. The new method consists of the numerical aggregation of infilling fines in a finite bed volume, during which the bed properties change to affect further infilling. An existing semi-empirical, particle packing model is implemented to identify these properties. It is shown that these types of models are adequate to describe regimes of natural sediment fabric quantitatively, and are thus useful tools in the analysis of sediment infiltration processes. Unlike previous developments to quantify saturated bed conditions, which assume that the supplied material is uniform and finer than the bed pore openings, the method developed herein considers poorly sorted fines, and can identify size fractions that are able to ingress into the bed due to being smaller than the particles that form the bed structure. Application of the new method to published experimental data showed that the final content of infiltrated fines is strongly sensitive to the initial bed packing density, highlighting the need to measure and understand open-work gravel deposits. In addition, the new method was shown to be suitable for assessing the degree of bed saturation, when it was applied to a published data set of field samples.