This paper proposes an alternative view on the allomorphy of the Dutch diminutive suffix –-je (with allomorphs: –-je, -–kje, -–pje, -–tje, and -–etje), an allomorphy allegedly determined phonologically. The major claim is that the whole of the allomorphy is not due to synchronic phonological processes. Moreover, the phonological process which does synchronically underlie the allomorphy is the strengthening of a palatal glide /j/. In particular, the -–tje and -–etje allomorphs do not derive from a /t/ through palatalization, but from a simple palatal glide /j/ which is strengthened in certain environments and is realized as [c]. Furthermore, allomorphs with epenthetic stops [p k] are not due to phonological alternation, but rather they may be argued to be lexicalized (relic) forms in fact. To account for the whole of the allomorphy, quite a number of restrictions should be postulated which would be operative on the diminutive only which points to morphological rather than phonological conditioning.