The literature pertaining to a variety of methods for estimating the abundances of songbirds is reviewed, and the use of two of the main methods - mapping and transects - is described and discussed. The mapping method is found to be the most efficient and reliable in the context of the present study, and work elsewhere in the British Isles suggests that the method has wide applicability in the region. Various methods for describing the habitats in which birds occur are discussed. In the present study the method chosen involved measuring various parameters of the vegetation within a limited area at each of several sampling points in five study plots in North-east Fife. It proved a useful method in the habitats studied - dune grassland, commercial and semi-natural coniferous woodlands, and deciduous scrub and mature woodland - and results of the vegetation sampling are discussed in terms of habitat structure and succession, and in the relationship of this structure to the breeding bird communities. The composition of these communities in the study plots is described, as are the changes over the three breeding seasons 1979-1981, which followed on from a meteorologically "hard" winter. Over the study period there was an overall increase in the populations of all five study plots, and this increase was greatest in the poorest habitat - the dune grassland - and least in the richest habitat - the mature deciduous woodland. Using indices to compare the study plots a major difference was found between the grassland plot and all four woodland plots, amongst which the two coniferous ones were most similar. Of the more abundant bird species, Robin, Song Thrush and Coal Tit were more prevalent in coniferous habitats; Willow Warbler and Blue Tit were more prevalent in the deciduous habitats. In relation to habitat structure, the more diverse bird communities were found in the more complex habitats, and there was evidence of succession in the bird communities concomitant with habitat succession. The present study is one of very few in Europe involving woodland habitats especially of a commercial nature near sea-level, and it was found that the four woodland study plots were rich in terms of the diversities and densities of their passerine breeding populations when compared with other European studies.