Simple ex vivo or in vitro models are most useful for testing putative cell therapy protocols, as they allow quick and controlled screening of variants and possible improvements. We discuss here three different models: coculture of precursors of human bone marrow cells (BMCs) with mouse heart slices bearing a cryogenic lesion; coculture of human BMCs and rat cardiomyocytes separated by a porous membrane that allows passage of soluble substances but prevents migration of nuclear material; and injection of human BMCs in developing chick heart bearing burn lesions. Our results indicate that the damaged areas express specific genes such as MPC1 and SDF1, and that some human BMCs migrate and graft near the lesion, where they can originate cells with a cardiac phenotype that produce human cardiac proteins. The frequency of this transformation is, however, very low. Understanding the factors that determine and regulate nuclear reprogramming and transdifferentiation would be crucial to appraising the contribution of these phenomena to cardiac regeneration and, eventually, to modulating them with therapeutic intent.