The ecological restoration of California's degraded and heavily invaded grasslands depends upon our ability to understand historical native species compositions. Since Clements' time, it has been assumed that, Nassella pulchra and other perennial bunchgrasses were the historically dominant species in California's pre-European grasslands. However, there has also been a lot of doubt. Relict analysis can be used to make historical floristic estimations. For this study, 13 relict grassland floras were surveyed and compared. A total of 1348 native grassland species were found. This means that these relict grasslands collectively served as habitat for about 40% of California's total native plant species richness. Many of the species occurred at just 1 or 2 sites and most species were annuals. Surprisingly, only 1% of the species were present in all 13 grasslands. This group consisted of 1 perennial herb, 10 annual forbs, 1 annual graminoid, and 1 perennial bunchgrass (N. pulchra). Jaccard distances were used to ordinate the grasslands in 2-dimensional space. Separation of grasslands along Axis 1 was strongly correlated with percentages of annual forbs and perennial grasses as well as mean annual precipitation. Latitude was most highly correlated with the distribution of grasslands along Axis 2. This study strongly suggests that, though N. pulchra was ubiquitous, grasslands of different regions of California had very broadly differing native species compositions. This indicates that we should take an explicitly regional approach to grassland restoration and that there is value in considering the entirety of community composition.