Conventional communication waveforms are composed of a small set of basis functions that are superposed to represent a stream of bits. Optimal detection of such signals in the presence of noise is achieved by a matched filter, i.e., a linear filter whose impulse response is the time-reverse of the basis function. In contrast, signals from well-known chaotic systems cannot be represented as a superposition of a small number of basis functions. Thus, no simple linear filter produces optimal detection through noise. Recently, a new family of hybrid chaotic systems has been identified whose solutions can be written in terms of basis functions. Since matched filtering can be applied to signals from these systems, their discovery suggests that chaotic waveforms may be suitable for carrying information in a practical communication technology. In this chapter, we first review the definition and significance of the matched filter in order to clarify the sense in which it is optimal. Next, we examine two hybrid chaotic systems in detail, and show how simple matched niters can be defined for their oscillations. Finally, we describe an experimental implementation of an electronic chaotic oscillator and its matched filter.