A single-stage modification of the turbine from a Mark 25 torpedo power plant was investigated to determine the performance with two nozzles and three rotor-blade designs. The performance was evaluated in terms of brake, rotor, and blade efficiencies at pressure ratios of 8, 15 (design), and 20. The blade efficiencies with the two nozzles are compared with those obtained with four other nozzles previously investigated with the same three rotor-blade designs. Blade efficiency with the cast nozzle of rectangular cross section (J) was higher than that with the circular reamed nozzle (K) at all speeds and pressure ratios with a rotor having a 0.45-inch 17 degree-inlet-angle blades. The efficiencies for both these nozzles were generally low compared with those of the four other nozzles previously investigated in combination with this rotor. At pressure ratios of 15 and 20, the blade efficiencies with nozzle K and the two rotors with 0.40-inch blades having different inlet angles were higher than with the four other nozzles, but the efficiency with nozzle J was generally low. Increasing the blade inlet angle from 17 degrees to 20 degrees had little effect on turbine performance, whereas changing the blade length from 0.40 to 0.45 inch had a marked effect. Although a slight correlation of efficiency with nozzle size was noted for the rotor with 0.45-inch 17 degree-inlet-angle blades, no such effect was discernible ,for the two rotors with 0.40-inch blades.Losses in the supersonic air stream resulting from the complex flow path in the small air passages are probably a large percentage of the total losses, and apparently the effects of changing nozzle size and shape within the limits investigated are of secondary importance.