DNA replication origins are the sites of assembly of replication-factor complexes that duplicate the genome, and they direct initiation in cells from all domains of life. Whereas bacterial chromosomes are replicated from a single origin, eukaryotes and most archaea use multiple origins and share similar replication machinery. Recent work by Allers and colleagues using the model halophile Haloferax volcanii suggests that this archaeal genome can also be replicated by an origin-independent mechanism. The circular main chromosome of wild-type H. volcanii has three origins (oris) that have different replication efficiencies. The replication profile of the H. volcanii laboratory strain showed a different pattern of ori utilization, which subsequent mapping revealed was due to chromosomal integration of the megaplasmid pHV4, resulting in four genomic oris. Remarkably, deletion of each ori singly and in combination to assess their relative activities revealed that cells lacking all four oris were still viable and had the same DNA content as did cells with a full complement of origins. Moreover, cells with no oris exhibited faster growth rates than did cells with all oris intact. Mapping of replication initiation sites in the ori-deleted strain identified no new discrete peaks in DNA copy number, which would have reflected activation of dormant origins. Instead, the ori-deleted strain showed a flat replication profile characteristic of dispersed initiation events. Genetic deletion analysis demonstrated that the ori-deleted cells are dependent on the radA gene encoding the archaeal homolog of bacterial RecA recombinase. Like RecA, RadA is essential for homologous recombination. Previous work has shown that Escherichia coli can use homologous recombination to initiate replication in the absence of a functional ori, but this mechanism severely compromises cell growth. The current work provides the first example of efficient duplication of an entire chromosome through recombination-dependent replication and raises intriguing questions concerning the efficiency and advantage of origin-directed replication to be addressed by future studies. (Nature doi:10.1038/nature12650, 3 November 2013)