The present paper is an attempt to empirically clarify certain aspects of the relationship between language and national or ethnic identity. We use the results of the sociolinguistic survey conducted in Northern Thailand in 2001. Two categories of variables are used: language variables represented by language ability and home language, and identity variables represented by self-name. The results of qualitative correlation/regression analyses in terms of phi coefficients and logistic regression show that; (1) language and self-name of a relative minority positively correlate with each other, as many previous studies argued, (2) a relative majority language, especially its use at home, negatively correlates mutually with the self-name of a relative minority. (3) the self-name of a relative majority is independent of, that is, neither positively nor negatively correlates with, the relative minority language.