Recently, the first investigation of nucleoli using mass spectrometry led to the identification of 271 proteins. This represents a rich resource for acomprehensive investigation of nucleolus evolution. We applied a protocol for theidentification of known and novel conserved protein domains of the nucleolus,resulting in the identification of 115 known and 91 novel domain profiles. Thephyletic distribution of nucleolar protein domains in a collection of completeproteomes of selected organisms from all domains of life confirms thearchaebacterial origin of the core machinery for ribosome maturation andassembly, but also reveals substantial eubacterial and eukaryotic contributionsto nucleolus evolution. We predict that, in different phases of nucleolusevolution, protein domains with different biochemical functions were recruited tothe nucleolus. We suggest a model for the late and continuous evolution of thenucleolus in early eukaryotes and argue against an endosymbiotic origin of thenucleolus and the nucleus. Supplementary material for this article can be foundon the BioEssays website athttp://www.interscience.wiley.com/jpages/0265-9247/suppmat/index.html.