We model the evolution of manganese relative to iron in the progenitor system of the globular cluster Omega Centauri by means of a self-consistent chemical evolution model. We use stellar yields that already reproduce the measurements of [Mn/Fe] versus [Fe/H] in Galactic field disc and halo stars, in Galactic bulge stars and in the Sagittarius dwarf spheroidal galaxy. We compare our model predictions to the Mn abundances measured in a sample of 10 red giant members and six subgiant members of ? Cen. The low values of [Mn/Fe] observed in a few, metal-rich stars of the sample cannot be explained in the framework of our standard, homogeneous chemical evolution model. Introducing cooling flows that selectively bring to the cluster core only the ejecta from specific categories of stars does not help to heal the disagreement with the observations. The capture of field stars does not offer a viable explanation either. The observed spread in the data and the lowest [Mn/Fe] values could, in principle, be understood if the system experienced inhomogeneous chemical evolution. Such an eventuality is qualitatively discussed in this paper. However, more measurements of Mn in ? Cen stars are needed to settle the issue of Mn evolution in this cluster.