Ewin (1976) was the first to describe two singing styles, "rapid'' and "slow', in the reed bunting, but he did not provide a satisfactory explanation for their occurrence. The present study demonstrates that the prevalence of song styles is related to male mating status. Songs of mated and unmated males differ mainly in temporal structure and in the composition of the syllables used. Since both singing styles are uttered at the same intensity, it is not likely that they are parts of one continuous scale, and it is justified to treat them as two different categories. Experimental removal of females shows that the prevalence of these singing styles is independent of season. Functional explanations in the context of mating and breeding behaviour are discussed.