Vocal learning and nonlearning birds frequently use the same habitats and are subject to similar selection pressures. However, although the repertoires of learning species are usually more complex, we know much less about how nonlearners encode information in acoustic signals. The present study describes the use of 3 basic mechanisms of acoustic encoding in signals of a vocal nonlearner: in repertoire, in acoustic structure, and in the temporal distribution of sounds. The study is based on observations of a simulated territorial intrusion in the Spotted Crake (Porzana porzana), a cryptic rail from Western and Central Eurasia. Males produced 5 types of calls. The loudest, the whitt call, consisted of 2 independent sounds with different fundamental frequencies: a soft and low F call and a loud and high G call. Before the playback, all males produced calls that consisted of both frequencies (F G calls), and most of the birds still produced such calls after the playback. Only birds that approached the speaker during the playback produced F calls, F G calls with a muffled G fundamental, and structurally distinctive soft rumble calls, whereas only birds that did not approach the speaker used G calls. These data suggest that males' engagement in aggressive interaction was associated with muting G or F fundamentals. The acoustic structure of whitt calls varied significantly between preplayback and postplayback recordings, and between approaching and nonapproaching males. However, certain acoustic parameters retained a high potential for individual recognition despite the playback. Finally, males lengthened their between-call intervals as they approached the speaker, which suggested that there was a link between the temporal distribution of calls and the aggressive motivation of males. Although they have a small repertoire of calls with innately programmed structures, Spotted Crakes modify their calling on many different planes and produce a high diversity of signals.