Girls in one subgroup of popular girls were selected by their teachers as having very positive academic behaviors ("popular studious" group). Girls in the other popular group, however, were rated by their teachers as having extreme antiacademic behaviors ("popular disengaged"). Three other groups were labeled "average popular," "unpopular disengaged," and "unpopular studious." "Popular studious" girls were rated by classmates as behaving in prosocial ways; whereas, "popular disengaged" girls engaged in many antisocial behaviors, often selected as "bullies" by some classmates. "Unpopular studious" girls were often labeled by their classmates as victims. Findings show a developmental character to the criteria for selecting popular peers. Elementary-school children generally selected friendly, hard-working peers; whereas, secondary-school children used contradictory criteria for popularity, with some picking friendly peers and other picking unfriendly, antisocial peers who scored low in academic performance. The uniformity of criteria for popularity increasingly diminished from ages 9 to 14. Participants were from 33 classrooms of 6 schools in 2 large Dutch cities. The 813 students in the study sample were close to representing the first-year cohort of students in each school. The homeroom teacher of each classroom completed a questionnaire on students' academic behaviors. Popularity scores were obtained by asking students to provide three names in response to the question, "Which students are the most popular in your class?" and three names in response to the question, "Which students are the least popular in your class?" Popularity based on friendship was assessed by having students indicate three classmates they considered their best friends and three classmates "with whom you would never be friends." 3 tables, 1 figure, and 56 references