The size and coding capacity of a genome are its most fundamental properties. What selective pressures deter mine the expansion and contraction of genomes, and how is genome size related to adaptive strategies? Very small marine microbial genomes are providing new insights into these questions. The first reports of genome sequences from the cyanobacterium Prochlorococcus and the α-proteo bacterium SAR11 (Pelagibacter) established that these very abundant marine bacterioplankton clades have unusually small genomes. The genome of Pelagibacter (1.31 Mbp) is the smallest reported genome for a free living heterotrophic cell. Prochlorococcus genomes, which range in size from 1.66 to 2.41 Mpb, are the smallest cyano bacterial genomes reported. Recently, the complete genome sequence of HTCC2181, an obligate methylotroph that is the first cultured member of the OM43 clade of marine bacterioplankton, was determined to be even smaller - 1.30 Mbp. These genomes are the smallest known from free-living cells, but larger than the genomes of some symbionts and parasites that live in close association with eukaryotic hosts. Genome streamlining has been invoked to explain the small genomes of some marine bacterioplankton. The essence of this theory is that selection is most efficient in microbial populations with large effective population sizes, causing the elimination of unnecessary DNA from genomes. Bacterioplankton may be particularly subject to streamlining selection and genome reduction because: 1) they have very large population sizes, 2) the live in a habitat that is frequently limited for the macronutrients N and P, which are stoichiometrically high in nucleic acids, and 3) selection favors high surface to volume ratios in the ocean surface, an adaptation that allows cells to compete effectively for nutrients. Recently, extraordinary examples of genome reduction in SAR11 have emerged. These cells are deficient in as similatory sulphate reduction genes, making them dependent on exogenous sources of reduced sulphur, such as 3-dimethyl-sulphoniopropionate (DMSP) or methionine, for growth. We also found that SAR11 cells are glycine auxotrophs that use a unique and elegantly simple glycine-activated riboswitch on malate synthase to control the assimilation of carbon through the TCA cycle into biomass. Studies of ultrastructure and the metaproteomics of cells from an extremely oligotrophic gyre show that these cells have high surface-to-volume ratios and very high ratios of transport proteins, an apparent adaptation to enable efficient replication in ocean “deserts”. These observations support the broad conclusion that metabolic versatility has been sacrificed for simplicity and genome reduction in some bacterioplankton, rendering them able to use ambient nutrient resources efficiently but reducing their versatility. The question remains, how does the evolutionary history and ecology of these organisms differ from microbial plankton with genomes of average size?