-We described geographic variation in the syllabic structure of the song of White-crowned Sparrows (Zonotrichia leucophrys nuttalli) resident in coastal California. We identified six large and relatively homogeneous populations that differed discretely at one locus in the song, the complex syllable, and have called these populations. Two superdialects were also identified on the basis of two other loci in the song, the introduction and the ending. Using another part of the song, the simple syllable locus, we also identified large-scale geographic clusters of similarity that were moderately concordant with dialect variation. From the patterns of geographic variation shown by the four song loci, we hypothesize that our study area was colonized by two ancestral populations. The existing song populations are therefore interpreted to be in secondary contact. Relatively narrow zones of overlap at dialect borders are identified and their widths estimated. Application of cline theory to the data suggests that 100-200 years have elapsed since secondary contact. One can approach the study of local geographic variation in cultural patterns, such as avian song dialects, in the same way that an evolutionary biologist approaches the study of geographic variation in morphological characters or gene frequencies (Cavalli-Sforza and Feldman 1981). An important difference is that, in many bird species, variation in song results from learning, not from genetic transmission. Recent surveys indicate that cultural inheritance of song is the rule for most oscine birds-nearly half of the world's avian species (Kroodsma and Baylis 1982). In this report, we apply a conceptual framework from evolutionary biology to interpret the spatial distributions of song variants in a population of birds. This application is appropriate because we hypothesize that in songbirds, as in humans, culturally transmitted traits affect the course of genetic evolution (Wyles et al. 1983). For long-term research on the biological significance of song variation, it is important to obtain a comprehensive description of the microgeographic variation in song among a series of contiguous populations. Such information provides a basis for understanding existing patterns of genetic differentiation of the populations, behavioral responses of both sexes to field and laboratory experimental manipulations (such as playback of male song of different dialects), and patterns of dispersal within and between dialects (Baker 1975; Baker and Mewaldt 1978; Payne et al. 1981; Petrinovich and Patterson 1981, 1982; Baker et al. 1982; Baker 1983; Tomback et al. 1983). Moreover, one may be able to suggest explanations of how song populations are related to one another in an historical sense (Baptista 1975). We as ume that populations with similar vocal patterns are more closely related than those with different vocal patterns, and that vocalizations can therefore be used to recontruct patterns of colonization (Baptista 1975, Mundinger 1975, Baptista 1977, Baptista and King 1980). Observations of zones of intergradation and hybridization, geographic variation, and clines have provided important information for interpreting evolutionary processes (Mayr 1963, Endler 1977), and our appr ach to interpreting song variation derives from this tradition. By analyzing song variation within and among the dialects of White-crowned Sparrows (Zonotrichia leucophrys nuttalli) occupying the Point Reyes Nat onal Seashore, we have developed a hypothetical model of how the dialects were derived. Because these populations occupy natural habitat in relatively undisturbed conditions within the pr tected boundaries of a National Seashore, our description of the geography of song may be of value in understanding the dynamics of song variation, both in the immediate d distant future. Evaluation of changes in the trait should provide an informative windo for observing the process of cultural evolution i avian communication (Jenkins 1978; Slater and Ince 1979; Ince et al. 1980; Mundinger 1980, 1982; Payne et al. 1981).