The expression of information encoded in genomes is not error-free. Transcript-error rates are dramatically higher than DNA-level mutation rates, and despite their transient nature, the steady-state load of such errors imposes a burden on cellular performance. However, a broad perspective on the degree to which transcript-error rates are constrained by natural selection and diverge among lineages remains to be developed. Here, we present a genome-wide analysis of transcript-error rates across the Tree of Life, showing that the effects of such errors are most likely at least partially dominant, and possibly synergistic, such that larger cells with more transcripts experience larger error burdens. Despite having a much narrower phylogenetic range of variation than genomic mutation rates, transcript-error rates vary in a manner that is consistent with the drift-barrier hypothesis, previously postulated as an explanatory framework for genome mutation-rate evolution. Thus, the degree to which natural selection is capable of reducing transcript-error rates is a function of both the population-genetic and the cellular environment (effective population size, cell volume, proteome size, and average fitness effects of individual errors). The idea that transcript-error rates are adaptively reduced in genes with high expression finds little support in the data.