A blade design study for a horizontal-axis wind turbine has demonstrated the use of a single-point optimization algorithm in concert with a general performance prediction program. The subject of the study is a 3-bladed machine with a radius of 45 feet and a generator power limit of 400 kilowatts. Estimated annual energy capture was the primary basis of comparison for competing designs. Three types of blade geometry were considered: those incorporating both optimum chords and twists at all spanwise stations, those having optimum chords for a constant, non-zero pitch angle, and those having optimum chords for blades with zero pitch angle at all stations. It was found that the loss of optimal twist caused the predicted energy harvest to decrease by about 14%, while the best zero-pitch blade suffered a 38% loss of annual energy in comparison to the optimally twisted blade. Additionally, several geometric simplifications made to the optimized blades for purposes of construction ease and reduced cost were seen to have relatively little effect on the estimated energy capture. Finally, it was shown that the use of airfoil data "synthesized" to produce constant chordwise force coefficients following stall will cause erroneous results when used with blade geometry having non-zero pitch angles.