This paper reassesses Ramer's evidence in favor of a genetic relationship between the Coahuiltecan languages, arguing that, given the loans among languages of the area, the role of borrowing should be accounted for in studies of linguistic kinship regarding these languages.