cardiorespiratory fitness with these endpoints. We excluded studies of occupational activity only, case-control designs, and studies that did not provide sufficient information for drawing a dose-response curve. We also excluded several studies that had been superceded. This left sixteen studies on physical activity (1,012,809 person-years of follow-up) and seven studies on fitness (312,195 person-years of follow-up). The physical fitness and physical activity studies were compared for their influence on relative risk. Relative risk is a measurement frequently employed in epidemiological studies for comparing the disease risk across groups. One group is defined as the referent group, often the group having the greatest risk. The relative risk is the ratio of the risk of other groups relative to the reference group. In these analyses, we chose the lowest fitness or physical activity category as the referent group. Thus, a relative risk for the referent group is always 1.0 because it is the ratio of its risk to itself. A relative risk of 0.7 means that the risk of disease is 70% of the risk of disease in the least fit or active category (or a 30% risk reduction), and a relative risk of 0.5 means that, compared to the least fit category, the risk of the disease is only half that of the referent group. The figure above presents the relative risks of the seven fitness and sixteen physical activity studies. The vertical axis (Y axis) displays the relative risk and the horizontal axis (X axis) displays the percentile of the population. Along the X axis, 0% is the least active and least fit and 100% is the most active and most fit. The lighter graph shows that the risks of CHD or