In vitro evolution is a technique used to explore the functional capabilities of RNA and DNA in a rapid, systematic, and directed fashion. In essence, this process mimics Darwinian evolution, only that "survival-of-the-fittest" is carried out at the level of individual molecules as opposed to individual organisms. The starting point is a very large pool of synthetic molecules with varying sequence of the four nucleotide subunits of RNA or DNA. Standard molecular biology methods are used to subject the pool to the desired selective pressure, create multiple copies of the surviving sequences, and prepare subsequent pools for repeated rounds of the selection and amplification cycle.

Starting pools may contain as little as thousands, or as many as trillions of different RNA or DNA sequences, depending on the challenge of the experiment. Chemically synthesized DNA is used either directly for DNA experiments or for in vitro transcription to derive a population of RNA. Molecules contain regions of short defined sequence that are necessary for molecular replication. The remainder of each molecule may either be a particular length of random (undefined) sequence or may be based on a particular sequence with mutations artificially introduced at a desired frequency. The starting pool then represents a collection of highly, or slightly, distinct sequences, respectively.