It has been recently claimed that the Inverse Amplitude Method provides a reliable unitarisation of Chiral Perturbation Theory allowing resonance poles to be accurately uncovered. We illustrate the sensitivity of these claims to the treatment of the Adler zero and to assumptions about the left hand cut (and hence about the underlying exchange forces). Previously favoured methods are shown to mistreat the Adler zeros and violate crossing symmetry casting doubt on the precision of their phenomenology. A more reliable solution is proposed. This places the Inverse Amplitude Method on a firmer basis.