Prior to European settlement, scattered, fire-tolerant trees, a dense herbaceous understory, and dynamic disturbance patterns, including fire and drought, characterized oak savannas. In the Midwest, only 0.02% remains, mostly as degraded, isolated fragments, and few studies have examined the post-settlement development of these oak savanna stands. In order to successfully restore and maintain oak savannas on today's landscape, we need to understand their functioning as controlled by biophysical conditions and disturbance patterns, which necessitates reliable identification of sites that previously supported savanna habitat. A common assumption is that the distribution of large open-grown trees reflects pre-European settlement structure; conversely, these trees may be the product of post-European settlement activities. We examined these alternatives by studying the composition and age structure of a remnant woodland site in central Iowa having structural characteristics indicative of past tallgrass oak savanna. Currently, the site supports large, open-grown Quercus alba L., in a dense stand of smaller Carya ovata (Mill) K. Koch, Ulmus americana L., Fraxinus pennsylvanica Marsh., and Ostrya virginiana (Mill) K. Koch. A total of 374 trees were analyzed for age using a combination of cross-sections from trees removed during a restoration treatment and cores from remaining trees. The oldest tree, Quercus alba, was 145 years old and does not predate settlement in the area. The majority of non-oak trees were less than 50 years old, coinciding with the time that grazing and cutting activities ceased. Our results suggest that the past cannot always be reliably inferred from the present, and that thorough analyses of stand development patterns in relation to historical land use change are needed to avoid incorrect or contradictory interpretations and to ensure sound restoration and management decisions.