This paper considers the numbers of species and individuals of terrestrial snails in small plots (quadrats) of different sizes (0.0625 m², 0.25 m², 0.56 m²) in three types of treeless fen sites differing in mineral content; a good proxy for calcium content. It compares the results of considering only live individuals or of combining these with empty shells, a common practice in land snail faunistic studies. As expected, the numbers of both species and individuals increased with increasing mineral content and with sampling area, whether all shells or only live specimens were considered. In two of the three sizes of plots and all fen (mineral level) types there was a clear increase of species when empty shells were included. However, the pattern of the increase varied among the fen types. The greatest increase was observed at the smallest plot size in mineral-rich fens; the difference was less in the larger plots. By contrast, in calcium-poor sites the increase was lower and the numbers of species did not change at the largest plot-size. This differential preservation affects interpretation. In very calcium-rich sites, empty shells contribute significantly to a summary of the fauna in the plot over several years, and can improve the inventory efficiency of sampling, provided that very old or subfossil shells of species no longer living in the site can be excluded. However, when comparisons are made among sites with different soil chemistry, estimates of densities and relative abundances will be distorted, because shells decay at different rates among them. Hence, in any studies concerned with densities or spatial heterogeneity within sites it is necessary to distinguish between live individuals and empty shells.