A weighted biological sequence is a string in which a set of characters may appear at each position with respective probabilities of occurrence. We attempt to locate all the tandem repeats in a weighted sequence. A repeated substring is called a tandem repeat if each occurrence of the substring is directly adjacent to each other. By introducing the idea of equivalence classes in weighted sequences, we identify the tandem repeats of every possible length using an iterative partitioning technique. We also present the algorithm for recording the tandem repeats, and prove that the problem can be solved in O(n²) time.