Patescibacteria form a highly diverse and widespread superphylum of uncultured microorganisms representing a third of the global microbial diversity. Most of our knowledge on Patescibacteria putative physiology relies on metagenomic mining and metagenome-assembled genomes, but the in situ activities and the ecophysiology of these microorganisms have been rarely explored, leaving the role of Patescibacteria in ecosystems elusive. Using a genome-centric metatranscriptomic approach, we analyzed the diel and seasonal gene transcription profiles of 18 Patescibacteria populations in brackish microbial mats to test whether our understanding of Patescibacteria metabolism allows the extrapolation of their in situ activities. Although our results revealed a circadian cycle in Patescibacteria activities, a strong streamlined genetic expression characterized the Patescibacteria populations. This result has a major consequence for the extrapolation of their physiology and environmental function since most transcribed genes were uncharacterized, indicating that the ecophysiology of Patescibacteria cannot be yet reliably predicted from genomic data.