Exercise enhances bone growth and increases peak bone mass. The aim of this study was to determine whether or not 4 weeks of deconditioning after 8 weeks of exercise in growing rats would result in a decrease in bone gain or reverse the benefits of exercise. Fifty 4-week-old female Sprague-Dawley rats were randomized by a stratified weight method into 5 groups with 10 rats in each group: 8 weeks exercise (8EX), 8 weeks sedentary control (8S), 12 weeks exercise (12EX), 8 weeks exercise followed by 4 weeks sedentary (8EX4S), and 12 weeks sedentary control (12S). The exercise consisted of running on a treadmill with a 5 degrees slope at 24 m/minute for 1 h/day and 5 days/week. After each period of exercise, cancellous and cortical bone histomorphometry were performed on double fluorescent labeled 5-microm-thick sections of the proximal tibia and 40-microm-thick sections of the tibial shaft, respectively. Eight and 12 weeks of exercise resulted in a significant increase in the body weight and gastrocnemius muscle weight by two-way analysis of variance (ANOVA). The femoral wet weight (mg; mean +/- SD; 8EX, 781 +/- 45.1 vs. 8S, 713 +/- 40.5; p < 0.05; 12EX, 892 +/- 41.6 vs. 12S, 807 +/- 19.8; p < 0.05) was significantly higher in the exercise group than that in the respective control groups. The femoral wet weight and bone volume (BV) of the 8EX4S group (818 +/- 46.2 mg and 531 +/- 31.2 microl, respectively) were significantly lower than those of the 12EX group (p < 0.05) and did not differ significantly from those of the 12S groups. The cancellous BV was significantly higher in the 8EX and 12EX groups than that in the respective sedentary groups (p < 0.05). The cortical bone area of the tibial shaft was also significantly higher in the 12EX than that in the 12S group (p < 0.05). The increase in the cancellous BV or cortical bone area was caused by an increase in the mineral apposition rate (MAR), without a significant effect in the labeled perimeter. The bone formation rate (BFR; microm3/microm2 per day) in the cancellous bone (12EX, 27.9 +/- 7.74 vs. 12S, 15.4 +/- 4.56; p < 0.05) or periosteal surface (12EX, 127.6 +/- 27.7 vs. 12S, 79.5 +/- 18.6; p < 0.05) was significantly higher in the exercised groups than that in the respective control group (p < 0.05). Again, deconditioning resulted in a decrease in the cancellous BFR, BV, periosteal BFR, and cortical bone area to levels not significantly different from the 12S group. In conclusion, our findings showed that exercised growing rats, when deconditioned, lost the benefits gained through exercise and their bone parameters were reduced to levels not different from the sedentary control. Thus, continued exercise is required to maintain high bone mass.