Suspended sediment (SS), ranging from nano-scale particles to sand-sized sediments, is one of the most common contributors to water quality impairment globally. However, there is currently little scientific evidence as to what should be regarded as an appropriate SS regime for different freshwater ecosystems. In this article, we compare the SS regimes of ten systematically-selected contrasting reference-condition temperate river ecosystems that were observed through high-resolution monitoring between 2011 and 2013. The results indicate that mean SS concentrations vary spatially, between 3 and 29mgL−1. The observed mean SS concentrations were compared to predicted mean SS concentrations based on a model developed by Bilotta et al. (2012). Predictions were in the form of probability of membership to one of the five SS concentration ranges, predicted as a function of a number of the natural environmental characteristics associated with each river's catchment. This model predicted the correct or next closest SS range for all of the sites. Mean annual SS concentrations varied temporally in each river, by up to three-fold between a relatively dry year (2011–2012) and a relatively wet year (2012–2013). This inter-annual variability could be predicted reasonably well for all the sites except the River Rother, using the model described above, but with modified input data to take into account the mean annual temperature (°C) and total annual precipitation (mm) in the year for which the mean SS prediction is to be made. The findings highlight the need for water quality guidelines for SS to recognise natural spatial and temporal variations in SS within rivers. The findings also demonstrate the importance of the temporal resolution of SS sampling in determining assessments of compliance against water quality guidelines.