Grading systems matter more to the teaching and learning enterprise than many teachers may realize, as demonstrated in the author's experience of adopting a new one. Different systems emphasize different values such as excellence vs. perfection, achievement vs. talent, and second chances vs. partial credit. The author relates her experiment with specifications grading, an outcome-based, pass/fail, rubric-based, and contractual grading system, and demonstrates its promise. She then addresses three questions her experiment raised: "Should I grade at all and if so, toward what end?" "Exactly what am I grading when I grade?" and "Is there any way to lessen the sting of failure?"