NaNbO3 has long been regarded as the 'most complex' perovskite known, with at least seven different phases reported as a function of T, at least two of which have evaded all previous attempts at structure solution. In the light of our own recent work exploting the complementarity of neutron and X-ray diffraction and solid state NMR, we now have a much better understanding of the issues involved in this apparent complexity. In fact, most existing samples of NaNbO3 at RT occur as a mixture of two similar but distinct phases, in space groups Pbcm and P21ma. We propose that some of the earlier studies have been impeded by this phase co-existence (without realising it!). Now we are able to synthesise a guaranteed phase pure sample of the Pbcm polymorph we are in a position to map out in full and correct detail the naNbO3 phase diagram. All earlier anomalies will be clarified and rectified.