One student began having seizurelike attacks in mid August 2002. During the following few weeks, both the number of affected individuals and the overall frequency of the episodes increased steadily. By October, episodes occurred up to 4 times per school day. With 2 exceptions, each student had only 1 episode in a given day. The number of attacks per adolescent varied from 1, in 3 girls, to 30 or more in 1 student, and most of the attacks occurred in 4 of the students. By December, the frequency of these attacks began to decline; after a 2-week holiday break, during which there was limited student contact, only the index subject and 2 other students continued to have periodic attacks for several more months. Most of the attacks occurred at school, but several girls reported occasional episodes in other settings, including 2 whose initial attack occurred while in church. According to the teachers, the attacks were unlikely to occur during a classroom session and often developed between classes in the hallways, in the cafeteria, or during breaks when most of the students were in the school yard. They also believed that the episodes were less likely to occur on rainy days when the students remained inside. A review of the nurse’s notes tends to support this observation: about half of the episodes with a documented time of occurrence took place around lunchtime.