Background/Question/Methods Understanding the factors that determine and maintain species diversity is a central topic in community ecology. After initial abandonment of agricultural fields, the community is characterized by low diversity which increases over successional time, as species with low dispersal ability are able to colonize. It is often assumed that ultimately, plant communities reach a stable equilibrium diversity level. However, even within later successional stages, plant species diversity may fluctuate over time. This variability may be due to changes in resource availability, disturbance regime, and the regional species pool. We examined plant species diversity in an “old-field” grassland plant community at Cedar Creek Ecosystem Science Reserve in Minnesota. The data was collected from 1900 permanently marked 0.5 m2 plots located in 19 fields abandoned between 1927 and 1982. Plots were sampled six times (every 3 to 5 years) over a 23 year period. Based on the species diversity at the previous time period, and rates of immigration and extinction, we utilized island biogeography theory to predict species equilibrium at the plot level. Results/Conclusions Immigration and extinction rates both varied over time and showed a curvilinear relationship with the diversity present. Extinction rates varied four times more, in relation to diversity, than immigration rates. The actual diversity fluctuated between 6.1 and 7.3 species per 0.5 m2. In contrast, the predicted equilibrium species diversity fluctuated between 4.1 and 7.9 species. However, the direction of change for both actual and predicted diversity was the same over each time period. The predicted equilibrium species diversity varied by as much as 50% of the diversity from the previous sampling period, whereas the actual diversity changed by only 20%. Actual diversity only partly reflects recent processes that drive species diversity, and in large part reflects past ecological processes. This is likely to be caused by the longevity of individual plants even in this relatively dynamic grassland plant community. Secondly, the community is oversaturated at some time points and undersaturated at others. Times when the plant community is undersaturated create windows of opportunity for the establishment of new species. Thus the key aspect determining ecosystem invasibility may be species turnover and the variability of immigration and extinction rates over time, not simply overall community diversity.