Abstract: In order to solve the problem that uncertain circumstances sometimes can not supply precise numerical values to the truth and false membership degrees of vague set, this paper puts forward the linguistic label vague (LLV) set. The truth and false membership degrees of any generic element of the universe are linguistic labels whose meanings are fuzzy sets. This paper also defines the complement union, intersection, and containment of LLV sets. As its application, this paper constructs a LLV decision table, acquires LLV decision rules, measures the strength of each rule by the degree of containment or intersection, and sorts the strengths of rules by the score function of LLV value. The result shows that the LLV set sometimes is more realistic in representing the imprecise information than the vague set.