Two experiments were conducted on Santa Cruz Island from 1993-2000 to evaluate the effectiveness of prescribed burns for increasing the abundance (species richness and cover) of native species and decreasing the abundance of alien species in annual grasslands. Although the two experiments differed greatly in scale and fire prescription, there were consistent outcomes between each. The overriding patterns were: (1) fire affected species composition only in the first 1-2 years post-burn, (2) climate and topography had far greater influence on patterns of species composition than fire did; (3) species response to fire was highly individualistic, and responses by “guilds” to fires are in actuality explained by the responses of just 1-3 dominant species within that guild; (4) regardless of fire season and frequency, alien annual species dominated cover in all years. These patterns indicate that outcomes from burning in annual grasslands will vary from year to year and site to site based on rainfall patterns and variation in soil and topography. Fire in and of itself will have little long term effect on modifying species composition in annual grasslands. From a conservation perspective, fire in annual grasslands will generally be most effective at maintaining, but not increasing, native species richness and abundance. However, in grassland restoration programs where the goal is to shift species composition to a higher proportion of native species, an effective approach could be the integration of fire with other forms of management (e.g. seeding).