It is widely assumed that the constraints of timed essay exams will make it virtually impossible for students to engage in the major hallmarks of the writing process, especially revision, in testing situations. This paper presents the results of a study conducted at Washington State University in the Spring of 2008. The study examined the occurrence of prewriting and revision in 890 timed essay responses as well as the impact of writing process on student scores. It was found that both prewriting and revision occur more frequently in timed essays than was previously realized. While prewriting corresponded to higher scores, revision corresponded to lower scores. These results encourage composition scholars to reevaluate their assumptions about both the validity of timed writing exams and the efficacy of current practices in teaching the writing process. (Contains 13 figures and 7 tables.)