Noise emissions are detrimental to the take-up of wind turbine technology. This is important for small wind turbine technology, as many are located close to dwellings. This paper presents a method of reducing turbine noise whilst retaining power production performance of small wind turbines using the numerical optimisation technique called differential evolution. The differential evolution technique coupled an empirical noise prediction model with a blade element-momentum analysis for predicting power performance. For the small wind turbine rotor at a design tip speed ratio of 5.5, a 2 dBA reduction in the total sound pressure levels was possible for a 1% reduction in power coefficient. This noise reduction was attributed to an increase in angle of attack and Reynolds number. The noise minimisation was then combined with starting time minimisation to improve low wind speed performance. At a tip speed ratio of 5.5, the best trade offs for a maximum of 1% decrease in power coefficient were a simultaneous decrease in total sound pressure level and starting time of 4% (2 dBA) and 6% respectively, or a decrease in total sound pressure level and starting time of 1% (0.7 dBA) and 16% respectively.