In 1987, J. Daniels and B. T. Roach published an exhaustive multidisciplinary review of evidence permitting the domestication and the early evolution of sugarcane to be traced. We try here to synthesize the new data that have been produced since, and their contribution to the understanding of the global picture. It is now highly probable that sugarcane evolved from a specific lineage restricted to current genus Saccharum and independent from lineages that conducted to genera Miscanthus and Erianthus. The scenario established by E. W. Brandes in 1958 is very likely the right one: Noble cultivars (ie. Saccharum officinarum) arose from S. robustum in New Guinea. Humans then spread these cultigens over large distances. In mainland Asia, natural hybridization with S. spontaneum occurred, and gave rise to the North Indian (S. barben) and Chinese (S. sinense) cultivars. Relationships between S. spontaneum and S. robustum in situations of sympatry are still not well understood.