A Bloom filter is a space-efficient probabilistic data structure that is used in many domains including networking applications to test for set memberships. Such applications often require sending Bloom filters using messages. Consequently, it is important to minimize the size of the filters such that the storage, transmission, and processing costs are minimized. In this paper, we introduce a novel data structure, which we refer to as the Compacted Bloom Filter (CmBF) that improves performance, uses less storage, and provides the same functionality as a standard Bloom filter. CmBF is suitable for many networking applications including IP traceback, peer-to-peer networks, and DDoS attack filtering approaches that require the transfer of large amount of information, such as IP addresses, among nodes. Moreover, CmBF works well for endpoints with limited memory. However, unlike the standard Bloom filter, CmBF introduces false negatives. We formalize expressions representing the false negative and positive rates. We perform simulations that validate our theory and investigate different tradeoffs in CmBF.