With the aging of society a great deal of interest is being placed on the value of longitudinal data in evaluating physiological losses. We present data on test-one/test-two reliability and reproducibility for measures of training, bone density from a longitudinal study of master athletes. Fifty-two males (mean age at test $1=58.2{pm}9.8;years$) and thirty-two females ($54.4{pm}8.8;years$) were selected from the study population. Bone mineral density was determined using DEXA (Hologic 1500). The characteristics of the subjects are presented below as $means;{pm};S.D$. The data was imported into the Statistical Package for the Social Science (SPSS 9.0, Chicago, IL). Paired t-tests were performed between visit 1 and visit 2 in subjects. Pearson correlations were performed. The results of this study indicate the measures of training history, body mass and bone density are reasonably stable and reproducible. We conclude that body composition and bone density parameters are stable and reproducible over time in active older subjects. Physiologic measures in master athletes are fairly stable, and reproducible over time Longitudinal studies investigating age-related changes in master athletes need to be conducted on a time schedule of greater than two years.