1. Nerve responses evoked by tone pips of known frequency were recorded from the medulla of Bufo americanus, Rana pipiens, Rana clamitans, Pseudacris triseriata, and Hyla crucifer. 2. Nerve responses were obtained between 20 cps and 4000 cps in Bufo and Rana, and between 100 cps and 6000 cps in Pseudacris and Hyla. 3. In all species, the best frequency for nerve responses was 1000 cps. In Pseudacris and Hyla there was a second low threshold area at 2000-3000 cps. It is suggested that the lower frequency dip in the response curve may represent responses from the pars amphibiorum and that the higher frequency dip may represent responses from the pars basilaris. 4. The threshold at the best frequency in Rana is similar to the sensitivity of human hearing. 5. Three Pseudacris and one Hyla were induced to answer tone pips. 6. These answered pips of frequencies between 700 cps and 4600 cps. Since this conforms to the high frequency dip in the nerve response curve, it is suggested that the Hylids might have been hearing pips with the pars basilaris only. 7. Pseudacris answered pips of a great variety of durations and intervals. 8. The data suggest that responses of Pseudacris males to frog calls may be, to a great extent, non-specific. 9. Alternating of calls in duets is controlled by at least two factors. There is a constant latency between a call and the stimulus eliciting it. A stimulus following a call too closely inhibits calling. 10. The data obtained here and by others suggest that even for location of males by females, the repetitive nature and frequency of calls may be of only minor importance. Rather, trill rate is implicated as an important call parameter.