Although early work on the tettigoniid T large fiber suggested that it might mediate early-warning and escape behavior in katydids, the majority of research thereafter has focused on the ability of the T-cell to detect, localize and/or discriminate mate-calling song. Interestingly, T-cell responses to conspecific song are rarely examined for more than a few seconds, despite the fact that many katydids sing for minutes or hours at a time. In this paper, the second of a pair examining the physiology of the T-cell in Neoconocephalus ensiger, we recorded T-cell responses using longer-duration playbacks (3 min) of conspecific song (Katydid signal 30 ms syllables, 9–25 kHz bandwidth, 12–15 kHz peak frequency) and two types of bat-like ultrasound, a 10 ms, 80->30 kHz frequency-modulated sweep (Bat 10 signal) and a 30 ms, 80->30 kHz frequency-modulated sweep (Bat 30 signal). Spiking responses were distinctly biased towards the short-duration ultrasonic signal, with more spikes per pulse, at a shorter spike latency and at a higher instantaneous firing frequency to the Bat 10 signal than to the Katydid signal or, surprisingly, to the Bat 30 signal. The ability of the T-cell to encode the temporal pattern of the stimulus was particularly striking. Only for the predatory bat signals did T-cell spiking faithfully copy the stimulus; playbacks of conspecific song resulted in significantly weaker spiking responses, particularly in male katydids. The results demonstrate that responses from the T-cell alone may be sufficient for katydids to discriminate biologically relevant signals pertinent to the phonotactic behavior patterns involved in mate attraction and predator avoidance.