The paper offers a critical analysis of the proposal to dismantle the genetic unity of the so-called Nile-Nubian languages by positioning one of its former constituents, the Nobiin language, as the earliest oshoot from the Common Nubian stem. Combining straightforward lexicostatistical methodology with more scrupulous etymological analysis of the material, I argue that the evidence in favor of the hypothesis that Nobiin is the earliest offshoot may and, in fact, should rather be interpreted as evidence for a strong lexical substrate in Nobiin, accounting for its accelerated rate of change in comparison to the closely related Kenuzi–Dongolawi (Mattokki–Andaandi) cluster.