In this paper, the feature selection algorithm with multi-granulation is proposed for symbolic interval-value data. A subset of a data set can be considered as a small granularity. Given a large-scale data set, the algorithm first selects different small granularities and then estimate on each small granularity the reduct of the original data set. Furthemore, it will introduce IT2 into hybrid fuzzy-rough QuickReduct algorithm to deal with interval-value data. Consequently, the algorithm will includes lower and upper dependency functions, lower and upper uncertainty degrees simultaneous. The “weighted” concept is not only enhance prior knowledge (important attribute), and also reduce unknown information effect. Fusing all of the estimates on small granularities together, the algorithm can get an approximate reduct.