To examine recent trends in physical education (PE) enrollment, daily attendance in PE, and being physically active in PE among high school students in the United States, this study analyzed data from the 1991, 1993, 1995, and 1997 national school-based Youth Risk Behavior Surveys (n = 55,734). Logistic regression analyses were conducted to test for significant linear time trends among the total student population and demographic subgroups (gender, race/ethnicity, and grade). Although PE enrollment in the total student population did not change from 1991 (48.9%) to 1997 (48.8%), the prevalence of students who attended PE daily, and the prevalence of students who were physically active > 20 minutes in an average PE class both decreased significantly among nearly all demographic subgroups. The prevalence of students who were physically active > 20 minutes in daily PE classes decreased from 34.2% in 1991 to 21.7% in 1997 (p < 0.001). To reverse current trends, high schools should implement daily PE classes that emphasize participation in lifelong health-related physical activity for all students.