Although researchers and engineers originally focused on a preponderantly irreversible computing paradigm, alternative models receive more and more attention. Reversible computation is a promising example which has applications in many emerging technologies such as quantum computation or alternative directions for low-power design. Accordingly, the design of reversible circuits has become an intensely studied research area. In particular, the efficient synthesis of complex reversible circuits poses an important and difficult research question. Most of the solutions proposed thus far are based on pure Boolean function representations such as truth tables or decision diagrams. In this paper, we provide a comprehensive introduction to and present extensions for the hardware description language SyReC which allows for the specification and automatic synthesis of reversible circuits. Besides a detailed presentation of the language׳s concepts and operations, we additionally propose algorithms that optimize the resulting circuits with respect to different objectives. A case study on a RISC CPU as well as a thorough experimental evaluation of both, the synthesis approach and its optimizations, show the applicability and demonstrate the advantage of SyReC compared to other solutions based on Boolean function representations.