The vegetative history of eastern North Da kota is broadly known from pollen and multi-proxy investigations of lake basins from across the re gion (Clark et al. 2001; Dean and Schwalb 2000; Wright et al. 2004). These records suggest that parkland or savanna rapidly developed following glacial retreat and subsequent drainage of glacial Lake Agassiz to the south (10,800 to 9400 14C year B.P.) with true grasslands (or prairie) emerging by 9000 14C year B.P. Changes in grassland com position of the region is poorly documented in the fossil record throughout the Holocene. While stable carbon isotopes offer some insight (Clark et al. 2001), this proxy is limited in that it cannot dis tinguish between changes in forest cover from C3 grasslands. Fossil phytolith assemblages are used to infer changes in Holocene grassland vegetation and cli mate at the Rustad site. The evidence presented comes from two stratigraphic columns allowing evaluation of local variability in phytolith assem blages and validation of results. The soils and sedi ments sampled in this study span much of the last 9,000 years. Several major environmental changes are hypothesized to have occurred during this time and should be detectable with opal phytolith analy sis. First, the early Holocene (12,000 to 8000 years B.P.) was a period of major climatic warming. Grassland response to this warming should show up in the phytolith record as a shift from a C3 grass dominated cover to a mixed grass prairie including both tall (mesic adapted) and short (xe ric adapted) C4 grasses along with diverse C3 grass taxa. Second, sometime during the mid-Holocene (8000 to 5000 years B.P.) many, but not all, re gional paleoenvironmental records suggest one or more periods of higher temperatures and greater aridity, broadly labeled the Altithermal. This mid Holocene aridity should be detectable in the phytolith record as a relative increase in phytolith morphotypes from the xeric-adapted, C4 short grasses such as blue grama, side oats grama, and buffalo grass. The goal of this chapter is to report the find ings of phytolith analysis and offer some prelimi nary interpretations. The analysis focuses on the grass phytolith portion of the assemblages, al though there is great potential for additional in formation from the arboreal phytolith component and from other proxy data, including carbon iso topes, terrestrial snails, and vertebrate fauna. METHODS