The considerable progress of recent years have demonstrated the potential of molecular electronics to first enhance the capabilities of silicon electronics and, possibly, to substitute it when the intrinsic physical limits of scalability will be reached by CMOS technology. Despite such advances, a real alternative will only be possible if novel circuit and architecture concepts will be developed, which are fully suited to molecular devices. The purpose of the present paper is to discuss the theoretical activities carried out at the Institute for Nanoelectronics of the Technische Universität München in the field of molecular devices and circuits, critically assessing such work with respect to what has been proposed in the literature. Two approaches to molecular electronics will be analyzed. On the one side, we will deal with transistors based on organic films, and present circuit solutions based on inverters as basic blocks. On the other side, we will consider single molecule devices and the architectures that might allow system implementations.