We present some key findings of a four-year, two-phase writing assessment project at Central Michigan University: Phase One (2002), a survey of faculty members (n=115) and subsequent focus groups (n=14) and Phase Two (2005), an evaluation of two samples of student writing (n=635 and 632). Major findings of Phase One reported here include the amounts and types of writing assigned by faculty members and their perceptions about the quality of their students' writing. Phase Two revealed some surprising results about our students' critical reading and writing abilities, confirmed the limitations of a timed-writing assessment methodology, and exposed an intriguing artifact of the data set. We reflect on the process of developing and conducting the assessment project, examine its strengths and weaknesses, and share our thoughts about the next phase of our assessment odyssey. (Contains 4 tables.)