During the early evolution of life, gene duplication, the production of two copies of a DNA sequence, allowed the rapid diversification of enzymatically catalyzed reactions and an increase in genome size, providing also material for the invention of new enzymatic properties and complex regulatory and developmental patterns. A duplication may involve (i) a part of a gene, (ii) a whole gene, (iii) DNA stretches including two or more genes involved in the same or in different metabolic pathways, (iv) entire operons, (v) a part of a chromosome, (vi) an entire chromosome, and finally (vii) the whole genome. Therefore, any DNA sequence may undergo a duplication event(s), but the fate of the replicate depends on whether it provides an evolutionary advantage to the host cell. Two hypotheses on the origin and evolution of metabolic pathways exist. The first one, the Horowitz retrograde hypothesis, predicts that an entire metabolic route was assembled by successive duplications of an ancestral gene in a backward fashion, starting with the synthesis of the final product, then the penultimate pathway intermediate, and so on down the pathway to the initial precursor. The patchwork hypothesis is based on the duplication(s) of ancestral gene(s) leading to the progressive increasing of specificity of low-specific enzymes, which then may be recruited to catalyze similar reactions in different metabolic pathways or sequential steps in the same route.