Most studies reporting cardiac changes with exercise have been cross sectional. The few available longitudinal studies have lacked standardization for environmental confounders. We prospectively assessed the relationship between increasing exercise intensity and cardiac remodeling in a highly standardized cohort of healthy young army soldiers.Sixty-three male army recruits (22 ± 3 yr) underwent a 12-wk moderate-intensity mixed strength and endurance exercise program, followed by a further 15-wk high-intensity exercise program, with highly controlled exercise, diet, and sleep patterns. Fitness (multistage fitness test), anthropometry, and 2D echocardiography were assessed.Moderate-intensity exercise was associated with increased fitness and decreased body fat % (both P < 0.01). There was no significant incremental change in these parameters after high-intensity exercise. By contrast, both moderate- and high-intensity exercises were associated with dose-dependent increases in left atrial and left ventricular (LV) volumes, LV mass, and right ventricular (RV) size (all P < 0.01). At the end of high-intensity training, 51% had a dilated LV and 59% had a dilated RV compared with published normal ranges. Almost all had normal LV systolic function and strain before and after exercise training. A small number of soldiers had mildly decreased RV systolic function at baseline and after moderate-intensity exercise (3% and 6%, respectively).We describe "soldiers' heart," which is characterized by balanced chamber dilatation, normal LV mass, and largely normal systolic function and myocardial strain. This prospective and highly controlled longitudinal study also found that increasing intensity exercise was associated with increasing chamber dimensions, which paralleled an increase in fitness after moderate-intensity exercise. After high-intensity exercise, however, cardiac chamber size continued to increase, but fitness did not increase further.