The study of gene expression is fundamental to systems biology, and reproducible and accurate measurements of relative and absolute protein concentrations are increasingly required for computational modelling and network-based analyses. In this work, we generated a library resource of peptide query parameters (PQPs) for the selective and reproducible quantitation of 2770 E. coli proteins, which we applied to quantify the absolute abundance of the proteome of E. coli across a broad spectrum of growth conditions using data-independent acquisition (DIA)-based SWATH mass spectrometry. A comparison of the obtained label-free peptide intensities with previously published protein synthesis rates measured by ribosome profiling allowed us to establish an optimized quantitative protein inference algorithm, termed xTop, and to determine protein mass fractions for thousands of E. coli proteins with concentrations as low as 50 copies per cell. We applied this workflow, which requires minimal investment in terms of sample amount and mass spectrometric measurement time (1 hour per sample), to study carbon and nitrogen starvation, translational limitation, as well as a variety of nutrient sources, growth conditions and stresses. This analysis allowed us to explore the hierarchical nature of regulation of the E. coli proteome across different scales: from “sectors” of proteins with similar response to nutritional or translational stresses, to clusters of proteins whose expression is strongly correlated across conditions, down to the detailed behaviour of individual proteins. In particular, we unveiled two large clusters of proteins associated to protein synthesis and general stress response.