Twenty speakers, diagnosed as male-to-female transsexuals, produced conversational recordings of speech and voice. The samples were submitted to perceptual evaluations and to acoustic analysis by means of a Visi-Pitch, Apple IIe microcomputer system. Transsexuals categorized as having female voices had higher fundamental frequencies (fo), less extensive downward intonations, a higher percentage of upward intonations and downward shifts, and a smaller percentage of level intonations and level shifts than transsexuals categorized as having male voices. The lowest average fo identified as belonging to a female speaker was 155 Hz. Higher (more feminine) ratings on the masculinity-femininity dimension correlated with fo (r = .89), percentage of level shifts (r = -.67), percentage of downward shifts (r = .50), percentage of level intonations (r = -.43), and percentage of upward intonations (r = .40). Findings are discussed in terms of the relative perceptual salience of average fundamental frequency and patterns of intonation for female voice quality.