Naval sonar has been accused of causing whale stranding by a mechanism which increases formation of tissue N(2) gas bubbles. Increased tissue and blood N(2) levels, and thereby increased risk of decompression sickness (DCS), is thought to result from changes in behavior or physiological responses during diving. Previous theoretical studies have used hypothetical sonar-induced changes in both behavior and physiology to model blood and tissue N(2) tension [Formula: see text], but this is the first attempt to estimate the changes during actual behavioral responses to sonar. We used an existing mathematical model to estimate blood and tissue N(2) tension [Formula: see text] from dive data recorded from sperm, killer, long-finned pilot, Blainville's beaked, and Cuvier's beaked whales before and during exposure to Low- (1-2?kHz) and Mid- (2-7?kHz) frequency active sonar. Our objectives were: (1) to determine if differences in dive behavior affects risk of bubble formation, and if (2) behavioral- or (3) physiological responses to sonar are plausible risk factors. Our results suggest that all species have natural high N(2) levels, with deep diving generally resulting in higher end-dive [Formula: see text] as compared with shallow diving. Sonar exposure caused some changes in dive behavior in both killer whales, pilot whales and beaked whales, but this did not lead to any increased risk of DCS. However, in three of eight exposure session with sperm whales, the animal changed to shallower diving, and in all these cases this seem to result in an increased risk of DCS, although risk was still within the normal risk range of this species. When a hypothetical removal of the normal dive response (bradycardia and peripheral vasoconstriction), was added to the behavioral response during model simulations, this led to an increased variance in the estimated end-dive N(2) levels, but no consistent change of risk. In conclusion, we cannot rule out the possibility that a combination of behavioral and physiological responses to sonar have the potential to alter the blood and tissue end-dive N(2) tension to levels which could cause DCS and formation of in vivo bubbles, but the actually observed behavioral responses of cetaceans to sonar in our study, do not imply any significantly increased risk of DCS.