School physical education programs are the principal provider of physical activity opportunities for most children and youth in the USA. A critical factor in the amount of physical activity students receive is their teachers' intentions to teach classes that are physically active. Understanding the determinants of teachers' intentions is an important first step toward increasing physical activity participation. This study used the theories of reasoned action, planned behavior and self-efficacy to examine the ability of attitude, social influences, perceived behavioral control, self-efficacy and teachers' own participation in leisure activities to predict perceived behavioral intentions of teachers to teach physically active lessons. The relationship between teachers' intentions and self-reported behaviors was also investigated. Participants were 399 (of those specifying gender, 225 female, 157 male), mostly Caucasian (85.7%) and African-American (7.3%), elementary, middle and high school physical education teachers. Participants completed a series of questionnaires grounded in social cognitive theory. Confidence in the validity of scores produced by the current instruments, which were slightly modified for physical education teachers, is based on the large body of research establishing validity of the instruments. The Godin Leisure-Time Exercise Questionnaire was also used to assess teachers' own leisure behaviors. The model was tested using hierarchical regression analysis. In block one, attitude and subjective norm were added to test the theory of reasoned action. Attitude and subjective norm jointly accounted for 46% of the variance in teachers' intentions to provide physically active classes. In blocks 2, 3, 4, the variables of perceived behavioral control, self-efficacy and teachers' reported leisure activity were entered, respectively. Only perceived behavioral control (4%) significantly (p<.05) contributed to predicting additional variance in intention. Teachers' intentions were also significantly correlated (R=.46) to a measure of how much time teachers spent with their students physically active at least 50% of class time. This study supports the significant effect of teachers' attitudes, social influences (i.e., students, parents, administrators, other teachers), and the control teachers feel they have over their curriculum, in influencing the physical activity provided to students. This study also supports the relationship between teachers' intentions to provide physically active classes and their actual self-reported teaching behaviors. This information can be used to help preservice and inservice programs address issues that may lead to increased physical activity levels in physical education programs.Keyword(s): physical activity, research, wellness/disease prevention