Adult-generated neurons in the dentate gyrus of the hippocampus have been the focus of many studies concerned with learning and memory (L&M). It has been shown that procedures like environmental enrichment (EE) or voluntary physical exercise (Vex) can increase neurogenesis (NG) and also enhance L&M. It is tempting to conclude that improvements in L&M are due to the increased NG, that is, a causal relationship exists between enhancement of NG and enhancement of L&M. However, it remains unclear whether the L&M enhancement observed after these treatments is causally dependent on the increase in newborn neurons in the dentate gyrus. It remains a possibility that some unspecified change–a third variable–brought about by EE and/or Vex could be a causal determinant of both NG and L&M. We suggest that this third variable could be neurotrophic and/or plasticity-related factors such as BDNF. Indeed, both EE and Vex can induce expression of such proteins, and BDNF in particular has long been linked with L&M. In addition, we argue that a very likely source of variation in previous experiments was the load on pattern separation, a process that keeps similar memories distinct, and in which NG has been shown to be critically involved. To attempt to bring these ideas together, we present preliminary evidence that BDNF is also required for pattern separation, which strengthens the case for BDNF as a candidate third variable. Other ways in which BDNF might be involved are also discussed.