Healthy premenopausal women were randomized into control and test groups; both exercised weekly in class and daily at home for a year. The test class did intermittent high-impact exercise; the control class did low-impact exercise. Bone density was assessed blind using dual energy X-ray absorptiometry at the femur (neck, Ward's triangle and trochanter) and at the lumbar spine (antero-posterior L1-4) on entry into the study, and again after 6 months (n = 27) and 12 months (n = 19). At 6 months the test group (n = 14) showed a significant increase of 3.4% in trochanteric bone density (p = 0.01) and this was significantly different from control (p = 0.05). In the second 6 months the control group was crossed over to high-impact exercise and showed a significant increase of 4.1% in trochanteric density (n = 7) while the original group maintained their improvement relative to baseline.