This paper considers the distribution of fieldwork locations for recent geographical research in Iceland, and explores the criteria on which fieldsite selection is based. Scientific considerations are not necessarily the dominant criteria in site selection. Where alternative sites are equally valid from a scientific point of view, the accessibility of sites, prior knowledge of sites by researchers, and personal associations between researchers are important factors in site selection. Middle-sized, sparsely developed islands such as Iceland are useful locations to explore issues of this kind, serving as 'natural laboratories' for the sociology and methodology of science as well as for science itself.