Tidal stream turbines are being designed and built to provide a renewable source of energy. The fatigue loads which are caused by varying environmental conditions and that contribute to the calculation of the design life of turbine components have been assessed. A quasi-steady loading analysis was used to compare differing unsteady inflow conditions with spectral and synthetic turbulence models, to determine damage equivalent loads (DEL) and was validated against published experimental data. It has been found that using a synthetic eddy method with the inclusion of blade scale fluctuations provides an efficient method to predict damage equivalent loads to within 5% of experimental values. However, a more computationally efficient method to predict the damage equivalent loads to within 7% is found by utilising the von KÃ¡rmÃ¡n turbulence spectra without the influence of blade scale fluctuations. Application of this method to long term environmental conditions based on a nine tidal constituent model of flow at the Pentland Firth shows that accounting for time variation of turbulence affects the DEL by a 20% decrease when compared to a mean constant turbulence value. Unsteady loading due to surface waves has also been analysed. Wave loading is determined using a time varying case of the Morison force equation, accounting for both waves and turbulence with following current. The Morison force is predicted to be within 11% of the experimental values of rotor thrust. The long term environmental rotor loading using a regular wave shows an increase by 1.74 over using an irregular wave spectrum. The DEL are used to compare the contribution of loads to the fatigue life. For the regular wave case the DEL are greater by a factor of 2.7 than the irregular wave case. Design conditions for a turbine in an array differ due to the onset wake generated by upstream turbines. The impact of this on unsteady loading and power has been assessed using both turbulence spectra methods and time-averaged upstream velocity wake deficits calculated using a Gaussian self-similar wake model. For turbines operating at constant tip-speed-ratio (TSR) to achieve close to peak power, the DEL can vary by up to 10-33%. In general it is found that by controlling or adjusting the TSR of the downstream turbines the DEL can either be made constant (0% variation) or the variation between turbines reduced (7-29%), compared to all turbines operating at the same TSR. This can be achieved with minimal impact on power, or in some cases this can increase the power. However, it does create slightly elevated net thrust; the impact of this on the onset flow velocity would need to be considered further.