All the classifications of present-day English irregular verbs to be found in the most famous grammars basically do not differ considerably and, especially, consider them to be mere aggregates of irregularities. This paper, on the other hand, aims to suggest a different classification by applying the basic tenets of Natural Morphology. This allows the author to divide irregular verbs into microclasses according to two main parameters: (i) the number of bases and (ii) rhymes. The latter is especially important, since empirical evidence demonstrates that the basic relationship existing among paradigms and families of paradigms is the one based on morphotactic similarity, whereas semantic similarities are absolutely peripheral. The two parameters are eventually exploited to provide a further scale, i. e. that of morphotactic transparency and of base uniformity, which is fundamental to organise a hierarchy of suppletion.