This study compared a three-year tracking of health-related physical fitness between two comparable samples of six graders that enrolled either in a PE4life middle school ("n"?=?154) or another school with a traditional PE program ("n?"=?93) in the United States. For the cohort, the FITNESSGRAM[TM] (Cooper Institute for Aerobics Research, 1992) battery was administered twice with a three-year interval, and moderate and vigorous physical activity levels in PE classes were assessed. Fitness tracking then was assessed with six statistical techniques and compared between the two samples. Students in the PE4life Program demonstrated significantly higher moderate and vigorous physical activity level ("p" < 0.001) in PE classes and improved significantly more in most fitness measures at follow-up ("p"s < 0.05 - 0.001) than those in the traditional program. Results of tracking assessed with Spearman correlations, Kappa statistics, and Kendall's tau-b did not reflect the significant differences in fitness changes between the two samples, whereas those assessed using stability in extreme quartiles, changes in percentage of students in at-risk groups, and odds ratios reflected the significant differences in the fitness changes between the two samples. The last three tracking methods identify differences in fitness changes in addition to assessing tracking stability and are recommended in fitness tracking research.