The ideal cell type to regenerate an acutely injured or chronically diseased lung would be a stem cell population from the patient's own lung. Consequently, extensive research efforts have focused on identifying and characterizing endogenous lung stem cells. Advances in techniques to facilitate cell isolation, labelling and tracking in vivo to determine their fate have led to the identification of several putative stem cell niches. Recently, convincing evidence has emerged for a novel stem/progenitor cell population in the submucous glands of the cartilaginous airways. These findings support the concept that there is no classical stem cell 'hierarchy' but that different progenitor populations within spatially distinct lung regions regenerate the lung epithelium adjacent to its niche. Intriguingly, recent findings challenge this concept; it was reported that the human lung may contain a primitive stem cell capable of differentiating into multiple cells of both endodermal and mesodermal lineage and of regenerating the injured lung. This suggests that a classical stem cell hierarchy may, in fact, exist in the lung. Although caution is needed in interpreting these emerging findings, the implications for our current concepts regarding lung stem cells, the insights into lung repair and regeneration, and the potential therapeutic implications are considerable.