Reconstructing the history of plastid acquisition and loss in the alveolate protists is a difficult problem because our knowledge of the distribution of plastids in extant lineages is incomplete due to the possible presence of cryptic, nonphotosynthetic plastids in several lineages. The discovery of the apicoplast in apicomplexan parasites has drawn attention to this problem and, more specifically, to the question of whether many other nonphotosynthetic lineages also contain cryptic plastids or are derived from plastid-containing ancestors. Oxyrrhis marina is one such organism: It is a heterotrophic, early-branching member of the dinoflagellate lineage for which there is no evidence of a plastid. To investigate the possibility that O. marina is derived from a photosynthetic ancestor, we have generated and analyzed a large-scale EST database and searched for evidence of plastid-derived genes. Here, we describe 8 genes whose phylogeny shows them to be derived from plastid-targeted homologues. These genes encode proteins from several pathways known to be localized in the plastids of other algae, including synthesis of tetrapyrroles, isoprenoids, and amino acids, as well as carbon metabolism and oxygen detoxification. The 5' end of 5 cDNAs were also characterized using cap-dependent or spliced leader-mediated reverse transcriptase-polymerase chain reaction, revealing that at least 4 of these genes have retained leaders that are similar in nature to the plastid-targeting signals of other secondary plastids, suggesting that these proteins may be targeted to a cryptic organelle. At least 2 genes do not encode such leaders, and their products may presently function in the cytosol. Altogether, the presence of plastid-derived genes in O. marina shows that its ancestors contained a plastid, and the pathways represented by the genes and presence of targeting signals on at least some of the genes further suggests that a relict organelle may still exist to fulfill plastid metabolic functions.