DNA sequencing of diatom assemblages from biofilms has already been used to assess the ecological status of freshwater in the UK. However, recent work using DNA data from these biofilms suggests that alternate metrics that capture the broader taxonomic and functional information to demonstrate importance of microbial biofilms could be useful. Exploring this potential requires large numbers of samples over time and space to be analysed. Sample archives could be used to meet this need, but the compositional stability of microbial communities in stored biofilm samples for more than one year is uncertain. This study compared changes in diatom assemblage structure using metabarcoding analysis of river biofilm samples before and after storage at -20°C in an RNAlater-based nucleic acid preservative. We found minimal changes in the diatom assemblages in the samples when stored for up to three years. Slight differences in certain groups observed were not significantly different suggesting any genuine differences in assemblages are masked by sub-sampling bias. These findings are similar to those observed in other studies looking at variations between analysts and sequencing instruments. This indicates that the microbial communities of the archived biofilm samples are stable. This will enable archived samples to be used for further research, including exploring microbial responses to environmental changes, potentially leading to the development of reliable microbial metrics and their integration into biomonitoring programs.