“The reconstruction of Neoarchaean and Proterozoic continents relies on comparing geological characters of the appropriate age between cratons. We present a numerical approach to analysing large igneous province (LIP) barcoding patterns based on parsimony algorithms that minimise inconsistency within the data by constructing minimal-length trees. These algorithms were developed in biology to uncover natural biological groupings and to describe the relationships between them, and are applicable to any data with a hierarchical structure that indicates ’common ancestry’. We present a proof-of-concept analysis using data from 370 to 100 Ma which recovered Pangaea, Laurasia, and Gondwana. Our results support the supercraton model in the late Neoarchaean/ Palaeoproterozoic (2.6–1.87 Ga), with the recovery of four craton groupings including Superia and Sclavia, with the implication that a modern tectonic regime had not yet developed. In the Proterozoic, two craton groupings for the age ranges 1.8–1.2 Ga and 1–0.56 Ga were recovered. The older time period produced two groupings: African, European, North American, and East Antarctica cratons in one clade and Australian, South American, North China and the Congo-São Francisco cratons in the other. Both Nena and Nuna are found as natural groupings within the former clade, but the overall arrangement does not conform to any published reconstruction of Columbia. The younger time period also produced two groupings (Laurussia and ‘west Gondwana’ plus ‘east Gondwana’) and does not support the idea of a single supercontinent of Rodinia.” In Precambrian Research