This article addresses the controversy about the reconstruction of the Proto-Indo-European consonant inventory, namely the question of whether the traditional, Neo-Grammarian reconstruction or one of the more recent alternatives such as the Glottalic Theory is correct. This question is directly related to that of the proper formulation of Grimm's Law. In the traditional framework the Proto-Germanic voiceless plosives derive from an Indo-European series of plain voiced plosives, whereas in the alternative account they derive from some kind of voiceless plosives, e.g. voiceless glottalics. The article brings to bear on the problem a new kind of evidence: the integration of prehistoric Semitic loanwords. This evidence unambiguously supports the traditional framework, in particular a shift of voiced to voiceless plosives under Grimm's Law.