From postrecording interviews of professional singers, it was hypothesized that recording environments, i.e., sound-treated environment versus an auditorium, may induce different vocal behaviors. To test this hypothesis, three groups consisting of nonsingers, singers, and actors were recorded in two different recording environments: a sound-treated booth (IAC) and an auditorium (AUD). Three recordings were obtained from each participant: recording one (IAC) and two (AUD1) required the participants to read in a normal voice; recording three (AUD2) required participants to pretend that they were "performing" before a full house. Results indicated that only the singers and the actors exhibited significant spectral and/or frequency/duration differences from one recording environment to another, with the most dramatic differences exhibited by the singers. It was concluded that the environment in which we record experimental samples from professional voice users, especially singers, should be considered as a variable that can affect results.