Recent high quality papers have renewed interest in the phenomenon of neurogenesis within the adult mammalian brain. Many studies now show that neurogenesis can be modulated by environmental factors including physical activity, stress, and learning. These findings have considerable implications for neuroscience in general, including the study of learning and memory, neural network plasticity, aging, neurodegeneration, and the recovery from brain injury. Although new light has been shed on this field, many contradictory findings have been reported. Here we propose two principle issues which underlie these inconsistencies, with particular focus on the interaction between learning and neurogenesis. The first issue relates to the basic methodology of measuring the generation of new brain cells, i.e., proliferation, as compared to survival of the newly made cells. Mostly, measures of neurogenesis reported are a combination of proliferation and survival, making it impossible to distinguish between these separate processes. The second aspect is in regards to the role of environmental factors which can affect both proliferation and survival independently. Especially the interaction between stress and learning is of importance since these might counteract each other in some circumstances. Reviewing the literature while taking these issues into account indicates that, in contrast to some findings, cell proliferation in the dentate gyrus of the hippocampus as a result of learning cannot be ruled out yet. On the other hand, increased survival of granule cells in the dentate gyrus as a result of hippocampal-dependent learning has been clearly demonstrated. Moreover, this learning-induced survival of granule cells, which were born before the actual learning experience, might provide a molecular mechanism for the 'use it or lose it' principle.