Genetic and migration data have provided inference on the factors affecting the 19th-century population structure of the Ǻland Islands, situated in the Baltic Sea between Sweden and Fin land. A genetic sample from the contemporary population was partitioned on the basis of grandparental origin in order to approximate genetic variation in the late 19th century. Bioassay of kinship, within and among 11 parish groups in these islands, provided genetic estimates which were compared with predictions of kinship derived from matrimonial migration patterns among these parishes in the period 1800–1849. The analyses used: (1) dendrograms describing hierarchical clustering of the groups; (2) a fit of the kinship coefficients to predictions under isolation by distance according to the theory of Malécot, and (3) two-dimensional representations of the kinship matrices rotated to best fit with the geographic locations of the parishes. Migration analyses suggested that the geographic distances among groups accounted for most of the variation in predicted kinship. Genetic estimates which incorporate more distantly determined founder effects and genetic drift as well as patterns of recent gene flow were much less related to the spatial pattern and showed that two parishes were rather distinctly isolated within the array. Differences between genetic and migration results could be attributed in part to problems inherent in the methodology. Overall, the analyses were in fairly good agreement and provided an interpretation of the factors affecting the Ǻland archipelago population structure in the past two centuries. Various aspects of the general problem of applying kinship theory to the analysis of population structure were also discussed.