The paper presents an algorithm for finding the measure of the union of intervals. The algorithm is based on, a divide-and-conquer approach. When implemented on a sequential computer it runs in optimal $O(n^*\log(n))$ time. The algorithm is modified for several parallel models of computations. The running time of the algorithm on mesh-connected parallel computers is optimal $O(n^{**}(l/2))$, while the running times on a concurrent read exclusive write PRAM and concurrent read concurrent write PRAM are $O(\log^{**}2(n))$ and $O(\log(n))$ respectively, which is very efficient.