This study investigates the half-lives and steady-state abundances of known and predicted *Escherichia coli* mRNAs at the single-gene level using two-color fluorescent DNA microarrays. The researchers developed an rRNA-based normalization strategy to quantify mRNA decay after transcriptional arrest by rifampicin. They found that mRNA half-lives were similar in nutrient-rich and defined media, with a wide range of stabilities observed for individual mRNAs, mostly between 3 and 8 minutes. Genes with biologically related metabolic functions often had similar stabilities. While a limited number of mRNAs showed positive correlations between half-lives and abundance, overall, increased mRNA stability did not predict increased abundance. The density of putative RNase E cleavage sites and the free energy of folding of 5′ or 3′ untranslated region sequences were not predictive of mRNA half-life. The results highlight previously unsuspected features of mRNA decay and suggest that generalizations about decay based on individual gene transcripts may have limited applicability.This study investigates the half-lives and steady-state abundances of known and predicted *Escherichia coli* mRNAs at the single-gene level using two-color fluorescent DNA microarrays. The researchers developed an rRNA-based normalization strategy to quantify mRNA decay after transcriptional arrest by rifampicin. They found that mRNA half-lives were similar in nutrient-rich and defined media, with a wide range of stabilities observed for individual mRNAs, mostly between 3 and 8 minutes. Genes with biologically related metabolic functions often had similar stabilities. While a limited number of mRNAs showed positive correlations between half-lives and abundance, overall, increased mRNA stability did not predict increased abundance. The density of putative RNase E cleavage sites and the free energy of folding of 5′ or 3′ untranslated region sequences were not predictive of mRNA half-life. The results highlight previously unsuspected features of mRNA decay and suggest that generalizations about decay based on individual gene transcripts may have limited applicability.