In many bird species the sex ratio of adults is male-biased, which is likely to have consequences for the ecology as well as for the conservation of a species. For example, when some males remain unpaired in a population, there should be strong selection on behavioural traits that enhance pairing success. A surplus of males is also likely to have important implications for the interpretation of breeding bird survey data. In our study population of Nightingales Luscinia megarhynchos, about half of the males stayed unpaired, suggesting that the number of males encountered singing was greater than the number of breeding pairs. Furthermore, the detectability (the probability of encountering a male singing) of mated males was only two-thirds that of unmated males when censused in the morning or late in the breeding season. The relative detectability was more similar early in the season and during the twilight periods before sunrise and after sunset. Males that arrived earlier on the breeding grounds were more successful in attracting a mate than males arriving later. Some of the unmated males deserted their territories and prospected areas up to 4000 m distant, whereas others settled on the study site only late in the season and may actually have changed territories. We suggest that adult sex ratios and the time of the census should be taken into account when interpreting the results of breeding bird surveys.