Programmed cell death is classified into apoptosis and autophagic cell death. The extensive crosstalk that occurs between these two types of death often prevents a clear identification of the leading death mechanism in a given experimental system. An accurate assessment of the type of death at work is of crucial relevance for the design of efficient cancer therapies aiming at eliminating tumor cells. Indeed, accumulating evidence indicates that resistance of tumor cells to apoptosis can be overcome by induction of autophagy. The latter would thus seem to represent an ideal strategy for eliminating certain tumor cells, except for the fact that autophagy induction may also contribute to cell survival. It therefore is of paramount importance to clarify the mechanistic links between autophagy and apoptosis as well as the nature of autophagy-dependent cell death. We recently reported that glioma cells resistant to death ligands were killed by the supernatant of activated microglia. What at first glance seemed to be apoptosis turned out to be autophagy-dependent cell death resulting from a blockade in the autophagic flux. This blockade most likely occurs at the level of lysosome recycling. We hypothesize that this autophagy-dependent process leads to either apoptosis or necrosis depending on the extent of lysosomal permeabilization and on the relative contribution of other cellular compartments. Autophagy therefore appears in our model as a cell-fate decision maker, not as a cell death execution pathway.