Purpose: The purpose of this study was to examine the predictive relationship among physical activity, health-related fitness, and on-task classroom behavior in children using a discriminant function analysis. Method: Participants were a convenience sample of children (N = 533; M[subscript age] = 8.8 ± 1.9 years) recruited from 77 1st- through 5th-grade classrooms at 3 low-income schools in a capital city in the Southwest United States. Percent of the school day spent in sedentary behavior (%SED), moderate-to-vigorous physical activity (%MVPA), and health-related fitness scores (body mass index [BMI] and Progressive Aerobic Cardiovascular Endurance Run [PACER] laps) were assessed during school hours. Classrooms were observed for on-task behavior during the academic year with the use of 5-s momentary time sampling methodology. A discriminant function analysis was performed using a binary on-task behavior outcome, stratified by an 80% on-task behavior cut point. Results: The results yielded 1 function (r[superscript 2] = 0.26, F = 13.1) explaining approximately one quarter of the total variance. The standardized function coefficients were -0.29, 0.29, -0.48, and 0.48 for %SED, %MVPA, BMI, and PACER laps, respectively. The sensitivity and specificity of the derived function for classifying a child into an on-task or off-task classroom were 0.79 and 0.73, respectively. Children who belonged to classrooms that achieved 80% on-task behavior displayed shorter times in sedentary behaviors (d = 1.01), lower BMI (d = 0.13), and higher PACER scores (d = 0.22) compared with children who belonged to off-task classrooms. Conclusion: School-day physical activity behaviors and health-related fitness scores can moderately discriminate children who belong to classrooms from low-income schools that are categorized as being sufficiently on task.