The classification of the nine world species of Pytho Latreille is reviewed by study of adult, larval, and pupal stages. Keys are provided for separation of species in these three life stages. Taxonomic changes (senior synonym in brackets) include synonymy of P. fallax Seidlitz 1916 [= P. niger Kirby 1837]; P. americanus Kirby 1837 [= P. planus (Olivier 1795)]; P. deplanatus Mannerheim 1843 is transferred from a junior subjective synonym of P. depressus (Linnaeus 1767) to a junior subjective synonym of P. planus (Olivier 1795). Lectotype designations are provided for the following: P. seidlitzi Blair 1925; P. nivalis Lewis 1888; P. niger Kirby 1837; P. fallax Seidlitz 1916; P. abieticola J. Sahlberg 1875; and P. americanus Kirby 1837. Eight larval stage, and 12 adult stage characters were selected for cladistic analysis. Lacking out-group material, pupal characters were not analysed. Character states were polarized using a generalized out-group composed of the three other genera of Pythinae (all monobasic). Phylogenetic analysis based on these 18 characters suggests four monophyletic species-groups: P. seidlitzi group (P. seidlitzi Blair — North America); P. kolwensis group (P. strictus LeConte – North America, P. kolwensis C. Sahlberg —Fennoscandia and the U.S.S.R., P. nivalis Lewis — Japan); P. niger group (P. niger Kirby — North America, P. abieticola J. Sahlberg — Europe, P. jezoensis Kôno — Japan); P. depressus group [P. planus (Olivier, 1795) — North America, P. depressus (Linnaeus, 1767) — Europe and the U.S.S.R.]. Larval stage synapomorphies are relatively more important in defining the species-groups than are those of the adult stage. The ancestor of Pythidae may have been associated with Coniferae as early as the Jurassic. The common ancestor of Northern Hemisphere Pythinae became isolated upon Laurasia once separation from Gondwanaland occurred near the end of the Jurassic. Two of the species-groups have similar disjunctions in North America, Europe, and Japan. The relatively eastern distributions of the North American member of each suggests that the ancestor of each species-group was Euramerican, and underwent vicariance with the opening of the North Atlantic in the Middle Cretaceous. The present distribution of both species-groups is thought to have been caused by the same vicariant event. The ancestor of the P. depressus group, which is presently circumboreal, was probably widespread and could have been Asiamerican in distribution. In the middle to late Tertiary, evidence suggests that Beringia was covered with coniferous forest, and the ancestor of the P. depressus group probably extended across this land bridge. Final separation between any North American and European/Asian species occurred in the Late Miocene or Pliocene, when a cooling climate made possible the evolution of treeless tundra in the north.