A UK seed conservation collection of Anemone nemorosa L. seeds held at the Millennium Seed Bank (MSB) showed low viability in its first post-storage test. Because achenes of A. nemorosa are naturally dispersed when they are green, we tested the hypothesis that seeds may not be fully desiccation tolerant and storable at the time of natural dispersal, and that a post-harvest treatment could increase the proportion of desiccation-tolerant seeds. Achenes harvested at the point of natural dispersal in late May in 2003 and 2004 were either placed immediately on 1% water agar at 20°C (‘laboratory’ treatment), or placed in nylon sachets and buried in leaf litter among plants growing in the wild (‘field’ treatment). Samples were withdrawn at intervals over a period of 168 d and tested for desiccation tolerance (drying to 0.059 g H2O (g DW)− 1) and longevity (controlled ageing at 60% relative humidity and 45°C). An initial increase, followed by a decline, in the proportion of seeds surviving desiccation and in the longevity of both laboratory- and field-treated samples coincided with the development of embryos from globular to heart- and then torpedo-shaped. Developmental arrest was not required for the acquisition of desiccation tolerance, and continued growth and development of the embryo resulted in a loss of desiccation tolerance, similar to that seen in orthodox seeds upon radicle emergence. Furthermore, while A. nemorosa seeds, like many from the Ranunculaceae family, might be described as having morphological or morphophysiological dormancy, this lack of developmental arrest does not fit with the usual concept of dormancy. The implications of these results for the classification systems of seed-storage behaviour and dormancy, and for the long-term conservation of seeds of A. nemorosa, are discussed.