In 2007 and 2008, controlled exposure experiments were performed in the Bahamas to study behavioral responses to simulated mid-frequency active sonar (MFA) by three groups of odontocetes: false killer whales, Pseudorca crassidens; short-finned pilot whales, Globicephala macrorhynchus; and melon-headed whales, Peponocephala electra. An individual in each group was tagged with a Dtag to record acoustic and movement data. During exposures, some individuals produced whistles that seemed similar to the experimental MFA stimulus. Statistical tests were thus applied to investigate whistle-MFA similarity and the relationship between whistle production rate and MFA reception time. For the false killer whale group, overall whistle rate and production rate of the most MFA-like whistles decreased with time since last MFA reception. Despite quite low whistle rates overall by the melon-headed whales, statistical results indicated minor transient silencing after each signal reception. There were no apparent relationships between pilot whale whistle rates and MFA sounds within the exposure period. This variability of responses suggests that changes in whistle production in response to acoustic stimuli depend not only on species and sound source, but also on the social, behavioral, or environmental contexts of exposure.