The flowfields of two axial-flow fans and one core compressor are studied in this thesis. The research data, previously gathered by other researchers, was obtained by means of a high-response dual hot wire aspirating probe. The probe measures total pressure and total temperature (and thereby isentropic efficiency) at several radial locations downstream of each rotor. The main objective of this research was to attain a heightened knowledge of sirbomachine losses and their source. To this end, two different methods of ensemble averaging the data are used in conjunction with data from the three rotors. Also, randomness calculations are performed for each rotor. The flowfield randomness of the two fans is shown to correlate well with their respective high tip losses. The core compressor is shown to have a very random flowfield, with no identifiable blade wakes, that extends across the entire blade span. This observation for the core compressor is supported by both contour plots and randomness calculations. Endwall boundary layer effects are thought to be much more predominant for the core compressor than for the two fans and are thereby the cause of the extent of the randomness in the core compressor flowfield. The randomness and losses for each rotor are expressed as a function of their inlet relative Mach numbers to see if shock losses are responsible for the randomness or losses. All three rotors are shown to have high tip losses. The high tip losses for the TS-30 rotor and Rotor 37 are shown not to be a function of inlet relative Mach number, while the TS-33 rotor has high tip losses that are thought to be partially due to a shock/boundary layer interaction. Aperiodicity, one category of randomness, is shown to increase with increasing inlet relative Mach number for the two axial-flow fans. Nonuniformity, another category of randomness, is shown to be independent of inlet relative Mach number for the two fans. / Master of Science