For many years CATV systems have been viewed as ideal for the implementation of two-way interactive systems. In the late seventies and early eighties, several revolutionary systems were developed which failed primarily because they were technology driven, rather than market driven. They attempted to address a wide range of interactive applications and became too costly and too complex to operate. Presently, two-way interactive systems concentrate on specific applications, such as Pay-Per-View and Home Shopping, and are developing in a more evolutionary manner. This paper summarizes the current state of the art and explores possible approaches for evolving current embryonic interactive systems into ones with more wide spread applications. Candidate control system architectures are analyzed.