The biggest unsolved problems in chloroplast evolution are the origins of dinoflagellate and euglenoid chloroplasts,which have envelopes of three membranes not two like plants and chromists, and of the sporozoan plastid, bounded by four smooth membranes. I review evidence that all three of these protozoan plastid types originated by secondary symbiogenesis from eukaryotic symbionts. Instead of separate symbiogenetic events, I argue that dinoflagellate and sporozoan plastids are directly related and that the common ancestor of dinoflagellates and Sporozoa was photosynthetic. I suggest that the last common ancestor of all Alveolata was photosynthetic and acquired its chlorophyll c-containing plastids in the same endosymbiogenetic event as those of Chromista. Chromistaand Alveolata are postulated to be a clade designated chrornalveolates. I propose that euglenoids obtained their plastids from the same(possibly ulvophycean) green alga as chlorarachneans and that Discicristata (Euglenozoa plus Percolozoa) and Cercozoa (the group including chlorarachneans) form a clade designated cabozoa (protozoa with chlorophyll a + b). If both theories are correct, there were only two secondary symbiogenetic events (witnessed by the chlorarachnean and cryptomonad nucleormorphs) in the history of life, not seven as commonly assumed. This greatly reduces the postulated number of independent origins of chloroplast protein-targeting machinery and of gene transfers from endosymbiont to host nuclei. I discuss the membrane and plastid losses and innovations in protein targeting implied by these theories, the comparative evidence for them, and their implications for eukaryote megaphylogeny. The principle of evolutionary conservatism leads to a novel theory for the function of periplastid vesicles in membrane biogenesis ofchlorarachneans and chromists and of the key steps in secondary symbiogenesis. Protozoan classification is also slightly revised by abandoning the probably polyphyletic infrakingdom Actinopoda, grouping Foraminifera and Radiolaria as a new infrakingdom Retaria,placing Heliozoa within a revised infrakingdom Sarcomastigota, establishing a new flagellate phylum Loukozoa for Jakobea plus Anaeromonadea within an emended subkingdom Eozoa, and ranking Archezoa as an infrakingdom within Eozoa.