There is a scarcity of marine mammal temporary threshold shift (TTS) data for long-duration, tonal noise exposures representative of naval continuous active sonar (CAS). TTS was measured in four bottlenose dolphins using psychophysical methods after exposure to frequency-modulated (FM) or pure tones with temporal characteristics simulating CAS. Exposures were conducted at multiple sound pressure levels and durations to test the equal-energy hypothesis. In phase one, the center frequency of the exposure was near the dolphins’ region of best hearing sensitivity (28 kHz). In phase two, the center frequency was 2.8 kHz, representative of actual frequencies of naval CAS. Larger TTS magnitudes were observed at 28 kHz for FM versus pure-tone exposures, indicating that the wider bandwidth did not reduce TTS effects. Additional electrophysiological measurements of both transient and steady-state auditory brainstem responses failed to show the consistent patterns of TTS observed using behavioral methods. For both exposure frequencies, the sound exposure levels corresponding to TTS onset were consistent with existing audiogram-based exposure weighting functions and the equal-energy hypothesis provided a good approximation of TTS growth. These data suggest that analyses of marine mammal TTS onset based on species-specific hearing sensitivity patterns and the equal-energy hypothesis are applicable to CAS exposures.