Circadian clocks are internal timekeepers that aid survival by allowing organisms, from photosynthetic cyanobacteria to humans, to anticipate predictable daily changes in the environment and make appropriate adjustments to their cellular biochemistry and behaviour. Whilst many of the molecular cogs and gears of circadian clocks are known, the complex interactions of clock components in time and space that generate a reliable internal measure of external time are still under investigation. Computational modelling has aided our understanding of the molecular mechanisms of circadian clocks, nevertheless it remains a major challenge to integrate the large number of clock components and their interactions into a single, comprehensive model that is able to account for the full breadth of clock properties. An important property of circadian clocks is their ability to maintain a constant period over a range of temperatures. Temperature compensation of circadian period is the least understood characteristic of circadian clocks. To investigate possible mechanisms underlying temperature compensation, I first constructed a comprehensive dynamic model of the Neurospora crassa circadian clock that incorporates its key components and their transcriptional and post-transcriptional regulation. The model is based on a compilation of published and new experimental data and incorporates facets of previously described Neurospora clock models. Light components were also incorporated into the model to test it and to reproduce our knowledge of light response of the clock. Also, experiments were carried out to investigate the unknown mechanisms of light response, such as the molecular mechanisms supporting the correct timing of conidiation after light to dark transfer. The model accounts for a wide range of clock characteristics including: a periodicity of 21.6 hours, persistent oscillation in constant conditions, resetting by brief light pulses, and entrainment to full photoperiods. Next, I carried out robustness tests and response coefficient analysis to identify components that strongly influence the period and amplitude of the molecular oscillations. These data measure the influence of the parameters in the model and were beneficial for making and testing predictions in the model. Thermodynamic properties were then introduced into reactions that experimental observations suggested might be temperature sensitive. This analysis indicated that temperature compensation can be achieved if nuclear localisation of a key clock component, FRQ, decreases with increasing temperature. Experiments have been carried out to validate this hypothesis and simulations were made to explore other possible mechanisms. However, from my experimental data and modelling results, the restriction of FRQ nuclear localisation might not be the only mechanism required to achieve temperature compensation. In conclusion, temperature compensation is most likely a complex property and may involve a combination of multiple mechanisms regulating clock component activity over a range of temperatures.