This thesis describes tests of a new hypothesis describing the effect of temperature on enzyme activity. Traditionally, the dependence of enzyme activity on temperature has been described by a model (the 'Classical Model') consisting of two processes: the catalytic reaction defined by ΔG‡cat, and irreversible inactivation defined by ΔG‡inact. To account for the anomalies found in the variation of enzyme activity with temperature, a new model (the 'Equilibrium Model') has been formulated to describe the effect of temperature on enzyme activity. In addition to the processes described by ΔG‡cat and ΔG‡inact, this model incorporates an inactive (but not denatured) form of the enzyme (Einact) that is in reversible equilibrium with the active form (Eact). The equilibrium between Eact and Einact is described by an equilibrium constant (Keq), whose temperature dependence is characterised in terms of the enthalpy of the equilibrium, ΔHeq, and a new thermal parameter, Teq, which is the temperature at which the concentrations of Eact and Einact are equal. This research has set out to: test the 'Equilibrium Model'; investigate the molecular basis of the temperature-dependent interconversion of the active and inactive forms of the enzyme; develop methods for the reliable determination of Teq, and outline the assay parameters required for accurate determination of Teq; examine the biotechnological implications of Teq; and finally, examine the evolutionary and ecological implications of Teq The 'Equilibrium Model' was tested by comparing 3D plots of experimental data (expressed as rate versus temperature versus time) collected for five enzymes with the 3D plots of the outputs predicted by the 'Classical' and 'Equilibrium' models. This analysis found that all five enzymes behaved as predicted by the 'Equilibrium Model', in displaying clear temperature optima at time zero, and led to the determination of plausible values for ΔG‡cat, ΔG‡inact, ΔHeq, and Teq. The value of Teq was affected when the enzyme-substrate interaction was altered (by the use of different substrates) but was, in general, unaffected by the addition of denaturing or stabilising agents to the assay. These results give some insight into the molecular basis of the equilibrium and, together with the fast timescale with which the Eact/Einact equilibrium occurs, support the hypothesis that it is unlikely that Einact is significantly unfolded and that the transition from Eact to Einact involves only a local (reversible) conformational change, possibly near or at the active site; in contrast to the slower and largely irreversible (under assay conditions) global conformational changes associated with thermal denaturation. The methodology for the determination of Teq was developed and extended to allow less laborious determination of Teq and to allow the determination of Teq using assay systems that are less than "ideal". Minimum assay parameters were determined in terms of sampling rate and temperature range. The potential implications of Teq for enzyme evolution, protein engineering and enzyme reactor performance have been introduced and discussed. The characterisation of three enzymes, once each from a psychrophilic, mesophilic and thermophilic source, has given some indication of the ecological and environmental implications of Teq, and suggests that Teq is a better indication of the source temperature of the enzyme than thermal stability and that ΔHeq reflects the thermal environment from which the enzyme was sourced. The results of this research are such that Teq can now be considered a fundamental thermal parameter of enzymes and is required alongside the Arrhenius activation energy and thermal stability to completely account for the way in which enzyme activity behaves with respect to temperature.