This report is addressed to administrators in the field of correspondence education. In dealing with attrition, it suggests that it is possible to control the proportion of those who do not start after enrollment; it may be possible to reclaim those who fail an examination; and the proportion of those who drop out may be reduced. Beginning with a survey of opinions in the field, the report examines correspondence study and attrition from an international perspective. A third chapter describes a way to compare attrition problems across institutions by distinguishing stages on a continuum from "career counseling" through "issuance of certification." The following chapters describe the Preliminary Unit Pack (PUP) experiment. Faced with high attrition and increasing enrollment, initial resources (often 300 pages of books and notes) sent to applicants on enrolling, as well as the staff needed to organize this process, were becoming expensive. PUP was begun to determine non-starters early in the program stages. The initial pack of resources sent to enrollees was reduced in size and a response to this material was required prior to formal enrollment. Nine of 100 courses were chosen for the PUP experiment. A reduction of the non-starter rate, with consequent savings of administrative resources, was achieved. A major side effect was that PUP students submitted assignments faster than others. Nine recommendations are suggested to reduce attrition in correspondence education (DAG)