Autophagy describes the process by which eukaryotes selectively and non-selectively target cytoplasm and entire organelles for lysosomal or (in yeast) vacuolar degradation. More than thirty different proteins contribute to this complex process, and it is widely recognised that the term autophagy does not describe merely a single linear pathway by which intracellular components are routed for lysosomal degradation. Yet, whilst autophagy has been unequivocally demonstrated in evolutionarily diverse organisms and the importance of autophagy in many aspects of human health and development is becoming ever more apparent, the extent to which autophagy in different taxa draws on a conserved cohort of readily recognisable proteins is not particularly clear. Here, we address this issue by comprehensive mapping of known autophagy components across a taxonomically diverse range of unicellular eukaryotes. Unexpectedly, our analysis points to independent examples of secondary loss of macroautophagy, the best understood of the autophagy pathways, in two parasites and one extremophile. Additionally, whilst our data point towards autophagy being an ancient innovation, utilising conserved core machinery, it is also clear that lineage-specific moderation (e.g. probable loss of Atg17 in some unikonts) and elaboration (paralogue expansion) of the core macroautophagy pathway occurs readily. Finally, we also consider the interplay between autophagy and organelle turnover in protists. Here, there are likely to be intriguing issues, as exemplified by mitochondrial turnover. In contrast to the dynamic mitochondrial fusion and fission observed in many eukaryotes (including yeast), cell-cycle regulated division of a single mitochondrion occurs in some protists. Yet, in these organisms mitochondrial function can often be rapidly re-modelled; we contend that in these species turnover of mitochondrial proteins is the product of intra-organellar protease activity.