Breakfast-eating frequency declines through adolescence and has been inversely associated with body weight in cross-sectional studies, with few prospective studies on this topic. This study was conducted to examine the association between breakfast frequency and 5-year body weight change in 2216 adolescents. Project EAT (Eating Among Teens) was a 5-year longitudinal study of eating patterns and weight concerns among adolescents. Surveys were completed in 1998-1999 (time 1) and 2003-2004 (time 2). Multivariable linear regression was used to examine the association between breakfast frequency and change in BMI, with adjustment for age, socioeconomic status, race, physical activity, time 1 BMI and breakfast category, and time 1 dietary and weight-related variables. At time 1, frequency of breakfast was directly associated with intake of carbohydrate and fiber, socioeconomic status, white race, and physical activity and inversely associated with smoking and alcohol consumption and dieting and weight-control behaviors. In cross-sectional analyses at times 1 and 2, inverse associations between breakfast frequency and BMI remained largely independent of all of the confounding and dietary factors. Weight-related factors (concerns, behaviors, and pressures) explained little of the breakfast-BMI association. In prospective analyses, frequency of breakfast was inversely associated with BMI in a dose-response manner. Further adjustment for confounding and dietary factors did not seem to explain the association, but adjustment for weight-related variables seemed to partly explain this finding. Although experimental studies are needed to verify whether the association between breakfast and body weight is of a causal nature, our findings support the importance of promoting regular breakfast consumption among adolescents. Future studies should further examine the role of breakfast habits among youth who are particularly concerned about their weight.