The performance of two constant-temperature normal hot-wire techniques for resolution of turbulent mass flux and local stagnation temperature in a supersonic flow is examined. The first technique used a single wire and the rapid scanning of multiple overheat ratios. Time averages of the signals at all overheats were used to separate the mean and rms mass flux, stagnation temperature, and their cross-correlation. The second technique used a dual-wire probe with each wire operated at different overheat ratios, giving instantaneous mass flux and stagnation temperature. In spite of a small separation distance (0.18 mm) between the wires in the dual-wire probe and high correlation between their signals, the rms mass flux inferred from the dual-wire technique was a factor of two higher than that from the single-wire technique. A consistency check based on data from one of the wires indicated that the dual-wire method produced results that were too high.