An RNA world is widely accepted as a probable stage in the early evolu
tion of life. Two implications are that proteins have gradually replac
ed RNA as the main biological catalysts and that RNA has not taken on
any major de novo catalytic function after the evolution of protein sy
nthesis, that is, there is an essentially irreversible series of steps
RNA --> RNP --> protein. This transition, as expected from a consider
ation of catalytic perfection, is essentially complete for reactions w
hen the substrates are small molecules. Based on these principles we d
erive criteria for identifying RNAs in modern organisms that are relie
s from the RNA world and then examine the function and phylogenetic di
stribution of RNA for such remnants of the RNA world. This allows an e
stimate of the minimum complexity of the last ribo-organism-the stage
just preceding the advent of genetically encoded protein synthesis. De
spite the constraints placed on its size by a low fidelity of replicat
ion (the Eigen limit), we conclude that the genome of this organism re
ached a considerable level of complexity that included several RNA-pro
cessing steps. It would include a large protoribosome with many smalle
r RNAs involved in its assembly, pre-tRNAs and tRNA processing, an abi
lity for recombination of RNA, some RNA editing, an ability to copy to
the end of each RNA strand, and some transport functions. It is harde
r to recognize specific metabolic reactions that must have existed but
synthetic and bio-energetic functions would be necessary. Overall, th
is requires that such an organism maintained a multiple copy, double-s
tranded linear RNA genome capable of recombination and splicing. The g
enome was most likely fragmented, allowing each ''chromosome'' to be r
eplicated with minimum error, that is, within the Eigen limit. The mod
el as developed serves as an outgroup to root the tree of life and is
an alternative to using sequence data for inferring properties of the
earliest cells.