To compare the effect of naval sonar up-sweeps and down-sweeps on the behavior of harbor porpoises, a harbor porpoise in a large pool was exposed to simulated low- and mid-frequency active sonar signals (series of 1-s duration frequency-modulated sweeps). Three sweep pairs were tested: (1) a 1 to 2 kHz up-sweep was compared with a 2 to 1 kHz down-sweep (both without harmonics) at a mean received sound pressure level (mean received SPL) of 114 dB re 1 µPa; (2) a 1 to 2 kHz up-sweep was compared with a 2 to 1 kHz down-sweep (both with harmonics; mean received SPL: 123 dB re 1 µPa); and (3) a 6 to 7 kHz up-sweep was compared with a 7 to 6 kHz down-sweep (both without harmonics; mean SPL: 107 dB re 1 µPa). For each sweep pair, the level was chosen during a pretest session with the intention that the harbor porpoise would respond to the sounds by moving away from the projector and surfacing more often (i.e., he would show a change in behavior). The study consists of three separate parts, so only a comparison within sweep pairs could be made and not between sweep pairs. For the 1 to 2 kHz sweeps with harmonics, the harbor porpoise swam further away from the sound source in response to the up-sweeps than to the down-sweeps. For the other two sweep pairs, sweep type (up-sweep or down-sweep) caused no significant difference in the harbor porpoise’s response. Thus, to allow the evaluation of potential effects of sonar sounds on harbor porpoises, sonar signal measurements should include the harmonics. For simulated naval sonar sounds with fundamental frequencies in the 1 to 2 kHz range containing harmonics, using down-sweeps appears to affect harbor porpoise behavior less than using up-sweeps.