Abstract BCL-2 family proteins form heterodimers or homo-oligomers to inhibit or induce apoptotic cell death, respectively. They often relocalize from the cytoplasm to mitochondria to carry out these functions. The traditional model is that in healthy cells, anti-death family members hold pro-death BCL-2 family members in check. Upon receiving a death stimulus, another set of proteins (BH3-only proteins) inactivate the protective BCL-2 proteins, forcing them to release their pro-death partners that are subsequently triggered to oligomerize and porate the mitochondrial outer membrane leading to cell death. In support of this traditional view, there is a preponderance of supporting evidence derived from the study of events that occur following treatment of cells with a death stimulus. Knockout and mutant mice also exhibit many developmental and treatment-induced phenotypes consistent with this model of antagonism between BCL-2 family proteins. Emphasis is logically placed on those phenotypes that support the model. However, this working model of BCL-2 family interactions has become so engrained that alternative, potentially valid interpretations are sometimes dismissed. Therefore, it is useful to consider the evidence that seems contrary to accepted models. In particular, the analysis of BCL-2 family functions in the nervous system has revealed unexpected outcomes that can serve to further stimulate critical probing of the yet unknown biochemical functions of BCL-2 proteins.