The study of energy landscapes holds the key to resolving some of the most important contemporary problems in chemical physics, such as how a protein folds to its native state, and why structural glasses exhibit a wide range of puzzling behaviour. For small molecules it is often possible to map out a complete reaction graph containing every permutational isomer and the transition states that link them. For water clusters, this approach has enabled us to predict and interpret tunnelling splittings observed in high resolution spectroscopy. For larger systems samples of local minima and transition states can be used to calculate thermodynamic and dynamic properties, and to visualise the landscape. The figures show representations of the potential energy surface for C60 (left) and a free energy surface for lysozyme (right). This energy landscape approach has recently enabled us to characterise the folding pathways and calculate rate constants for peptides, and applications to larger proteins are now underway.