This paper dramatizes the complexity and the problems involved in calculating the rates of student dropouts from school. To compare the dropout formulas used by various agencies, states, and local school systems, responses from a national survey are presented and used to calculate a range of dropout rates for the Austin (Texas) public schools. By using the membership, enrollment, and dropout counts of a single district, it is possible to demonstrate the differences among the contrasting dropout rate formulas being used around the country. These formulas (including single-year rates, multiple-year rates, graduation rates, and census-based rates), when applied to Austin, produce dropout rates ranging from 10.1 percent to 50.6 percent. Other considerations affecting the reported dropout rate include time segmentation, transfer students, student mobility, and changing demographics. A true graduation rate has better possibilities for communicating the success of a school system than a dropout rate. The most representative and valid single-year dropout rate would be characterized by: (1) a 12-month calculation period; (2) a cumulative count of students; and (3) sufficient time to allow students to be documented as transfers. Also proposed is a 4-year longitudinal rate using a cohort of first-time ninth graders. (23 references) (MLH)