The purpose of this study was to determine the characteristics of effective internships in principal preparation programs. A three-round modified policy Delphi was used. The panel consisted of interns, mentors, and university facilitators from 18 of the 22 Danforth Foundation-funded principal preparation programs. Characteristics of internships were divided into two categories: procedural and experiential. Procedural characteristics were defined as the process and management characteristics of the internship. The procedural characteristics were subdivided into selection, planning, placement, support, and evaluation. Experiential characteristics were defined as the experiences the intern had once placed with the supervising administrator. The experiential characteristics were subdivided into school climate; instructional management; research, evaluation, and planning; resources; staff development; staff evaluation; and community support. From the data base for Round 1, 132 items were selected for subsequent rounds. Panel members were asked to rate the importance of each item on a five-point scale. Items with a mean of 3.75 or greater were considered a characteristic of an effective internship in a principal preparation program. At the completion of Round 3, 114 out of 132 items met this criterion. Consensus by interns, mentors, and university facilitators was reached on 116 out of 132 items. Consensus was defined as no statistically significant difference among the ratings of the three groups. The major conclusions drawn from the study were (a) there are a number of different ways that an internship may be implemented in a principal preparation program, (b) the intern should not have full-time teaching (or other job) responsibilities while completing the internship, and (c) educators are somewhat parochial when it comes to recognizing the benefit of experience outside of education. It was recommended that the characteristics identified in the study be reviewed by those who develop internships in principal preparation programs and that the procedural and experiential characteristics that fit the context of the program be considered for implementation. / Ed. D.