Uracil, a close thymine analog, is traditionally considered as a mistake to be corrected in DNA. Repair of uracil as a spontaneous cytosine deamination product is essential, but the repair enzyme uracil-DNA glycosylase usually also eliminates thymine-replacing, "innocent" uracils, as well. Inhibition of dUTPase and/or thymidylate synthase results in elevated cellular dUTP/dTTP ratio that, due to the suboptimal specificity of DNA polymerases, leads to incorporation of thymine-replacing uracils to an extent that overloads the repair apparatus and induces apoptosis (thymine-less cell death). This research field is of interest for therapeutical applications, as well. Novel, partially isolated and yet controversial, experimental data, together with results from our laboratory argue for a re-interpretation of the traditional view on uracil being solely a mistake in DNA. Several studies indicate that uracil-DNA may be at least transiently tolerated in different organisms, and it may possess physiological or developmental role. We have reviewed such examples and the novel results that reinforce the hypothesis on uracil-DNA being a developmental death signal in the pupal stage the fruit fly (Drosophila melanogaster), due to lack of dUTPase and uracil-DNA glycosylase in the fruit fly larvae. Elevated uracil content in DNA of the larvae and the existence of a uracil-DNA degrading nuclease under strict developmental control at pupal stage present key pieces of experimental evidence in this hypothesis.