Interval-Valued Fuzzy Sets handle uncertainty and vagueness eectively. These features are particularly useful for clustering. In this paper it is showed the utility of Interval-Valued Fuzzy Sets for clustering with no accurate information. An easy method for clustering is proposed by generating transitive closures under a pseudo-t-representable t-norm. Clusters are computed from transitive closures by generating alpha-cuts. It is found that some of these alpha-cuts are equivalence classes under the pseudo-t-representable min. It is also found that these transitive closures are closer to the original interval-valued fuzzy relation that the classical transitive closure under the t-norm [min,min]