Quantifying maximum stand density is important to evaluate the potential stand density of the target population in forest management. However, most of the past research in the eastern US was mainly focused on planted monocultures, coniferous forests, a few commercially important species at the stand level, or in a particular geographic region. This study aimed to estimate the maximum stand density for mixed-hardwood forests across physiographic zones in the eastern US between two decades (1996–2009 and 2010–2021). Data used in analyses were collected from the US national forest inventory established and maintained by the USDA Forest Service’s Forest Inventory and Analysis (FIA) program.Results showed that the slope of the self-thinning lines varied among forest types. Estimating maximum stand density index (SDImax) from size-density relationships produced more precise estimates than using SDI-size curves. Among all forest types, elm-ash-cottonwood (Ulmus-Fraxinus-Populus) showed consistent SDImax estimates whereas other forest types varied by regions. New England had considerably higher SDImax in aspen-birch (Populus-Betula), oak-hickory (Quercus-Carya) and oak-pine (Quercus-Pinus) forests than other physiographic zones. Most of the combinations showed consistent SDImax between two time periods. Only six combinations showed a significant gain (4–14% increase), which was likely driven by the growth of the same dominant species groups. The findings of this work provided not only additional insights of maximum stand density in the region, but also a methodology for forest ecologists and managers to quantify SDImax for a variety of forest types.