The field of linguistics has no standard way of proving that two languages are related. Nevertheless, several approaches have been suggested for evaluating whether the similarities or correspondences offered in support of a relationship between two languages are characteristics that are highly unlikely to have arisen by chance. The central purpose of this chapter is to explore any statistical method that might provide confirmation or disconfirmation that the lexical similarities, grammatical similarities, and sound correspondences observed between Proto-Basque, as reconstructed here, and Proto-Indo-European as standardly reconstructed, are not due to chance. As a consequence, two distinct methods are considered. In the first part of the discussion, results of applying Nichols’ “individual-identifying threshold” to grammatical and lexical data from the two proto-languages, is put forward. Lower numerals, pronouns, derivational paradigm, striking homophone correspondence set, and several individual-identifying lexemes are offered as evidence. In the second part, Kessler’s (2001) ‘Monte Carlo’ method is discussed. For reasons specific to the Basque data, some revisions of this method are necessary. A method based on language-specific segment inventories and phonotactics may offer more reliable results. This technique has been applied successfully to Proto-Indo-European/Proto-Basque Swadesh lists as summarized in §8.2.