RNA world hypothesis states that RNA was, before the emergence of the first cell, the dominant, and probably the only, form of life.
This hypothesis is supported by RNA's ability to participate in the storage, transmission, and duplication of genetic information, similarly to DNA, coupled with it's ability to to act as a ribozyme (similar to an enzyme), catalyzing certain reactions.
In a RNA world, different forms of RNA compete with each other for free nucleotides, and are subjected to natural selection. The most efficient molecules of RNA, the ones able to catalyze their own reproduction, survived and originated the modern RNA. It's possible that competition between RNA favored the emergence of cooperation between RNA molecules opening the way to the formation of the first proto-cell.
At first glance, the RNA world hypothesis seems implausible given that in today's world large RNA molecules are especially fragile, subject to hydrolysis that degrades these long biopolymers into their constituent monomeric nucleotides. However, in today's world enzymes capable of catalyzing this hydrolysis called RNAses ("ar-en-ases") are ubiquitous, contributing to this fragility. In a pre-biotic world absent any protein, including RNAses, a given RNA molecule might have "lived" longer then than it can today.
The RNA world hypothesis, if true, has important implications for the very definition of life. We may be too ready to define life in terms of DNA and proteins. After all, in today's world, DNA and proteins seem to be the dominant macromolecules in the living cell, with RNA serving, for the most part, as a mere messenger between them. But the RNA world hypothesis places RNA at center-stage when life originated, and therefore requires that we define life primarily in terms of RNA, and only secondarily in terms of DNA and proteins. If the RNA world hypothesis is true, life can be defined as the set of strategies that RNA polynucleotides have used and continue to use to perpetuate themselves.