Background Oscillators are a fundamental building block in many engineering fields, as they provide the basis for counting, timing and synchronisation. Oscillators are found in many everyday devices such as clocks, computers or radios. Similarly, oscillations are an essential part of biological systems – providing the basis for, for example, rhythmic patterns and regulatory networks. The ability to build a stable, controllable biological oscillator would be a major step towards reliable synthetic biology based circuits. Elowitz et al. were part of the first ones to try to build an oscillator. Their oscillator was based on genetic network at the single cell level. However, due to the stochastic behaviour inherent at the gene expression level, the oscillations were not stable or persistent. In this paper, we present an original oscillator design produced during the iGEM-2006 competition at Imperial College. The project addressed stability and reliability issues by defining an oscillator at the population level and by applying strict engineering rules to the system development process.