The influences of beginning-level gymnastics training on skeletal development were assessed in prepubertal females four to eight years of age, with essentially no organized physical activity experience prior to the onset of training. The first study was conducted using a two-year prospective quasi-experimental design to examine the influence of participation in gymnastics on bone mineral accrual (Chapter 3). One hundred and ninety six female children were recruited based on the criteria that they had not participated in organized physical activity (or had limited participation <12 weeks). Approximately 80% completed the two-year study (n=156). Sixty-five of these individuals elected to enroll in recreational gymnastics classes and were compared to 78 controls participating in nongymnastic activities or no activities. At study initiation, children who elected to begin gymnastics training were significantly shorter, lighter and leaner than those choosing to participate in the control group. Furthermore, most measures of bone mineral were lower in gymnasts vs. controls at the study onset. Over two years, gymnasts experienced a significantly (p < 0.05) greater rate of increase in areal bone mineral density (aBMD) of the lumbar spine (3.5%) and bone area of the radius (3.6%) compared to controls. To more rigorously control for biological and maturational differences between groups, a subgroup of prepubertal gymnasts (n=31) was individually matched to a subgroup of controls (n=31) based on prepubertal development (Tanner stage I throughout the two-year investigation), race, age, height and weight. Similar to the observations in the overall sample, the gymnasts in this group had significantly (p < 0.05) higher rates of lumbar spine aBMD (2.7%) and total proximal femur aBMD (1.5%) accrual compared to the matched sample of controls. Those gymnasts who advanced to a higher-competition level (n=9) were compared to those who remained at a lower-noncompetitive level (n=56) and revealed that the higher level gymnasts gained more aBMD at the lumbar spine (3.9%) and radius (3.0%) compared to low-level gymnasts. Using data from the matched subgroups (n=31 per group), the study in Chapter 4 investigated the influence of gymnastics participation on geometric strength properties determined by hip structural analysis of the proximal femur (PF). No differences in structural properties existed at baseline, and over two years, gymnasts, compared to controls, had moderately greater increases in cross-sectional area, cross-sectional moment of inertia and section modulus at the narrow neck. These relationships depended on initial weight, where gymnasts who were heavier demonstrated the greatest strength benefits over controls. Conversely, endocortical thickness increased significantly more in the controls vs. the gymnasts. Controls also had greater increases in subperiosteal width compared to gymnasts, but this relationship depended on initial weight. At the intertrochanteric region, gymnasts had moderately greater increases in cross-sectional moment of inertia and section modulus compared to controls, whereas controls had greater increases in subperiosteal width. The interactions observed for changes in cross-sectional moment of inertia and section modulus both depended on initial weight, whereas the changes in subperiosteal width depended on initial height. Over two years, gymnasts did not differ from controls in strength variables at the shaft region of the PF. Higher-level gymnasts (n=9) showed no geometric differences over time in the PF compared to a matched group of low-level gymnasts (n=9). Findings from these studies suggest that the initial two years of recreational artistic gymnastics training in prepubertal females increased aBMD at the lumbar spine and bone area of the radius beyond those observed in controls. Furthermore, the increases in aBMD of the total proximal femur with gymnastics training observed in the matched prepubertal sample (n=31 per group) translated into moderate differences observed in the structural properties of the PF, as assessed by hip structural analysis. However, the improvements that were observed depended on initial height and weight, where those gymnasts who were taller, heavier or more developmentally mature, had the greatest structural advantage over controls. It may be possible that the positive effects of gymnastics participation on estimated bone strength in the PF will emerge as these young females advance in maturity.