A longitudinal experiment was designed to test the hypothesis that individual mice differ in their aging rate and to validate candidate biomarkers proposed to measure the rate of aging. Mice were bred as the genetically heterogeneous progeny of a cross between CB6F1 mothers and C3D2F1 fathers. Half of the mice were fed ad libitum (AL group), and the other half were subjected to 60% calorie restriction (CR group). Each mouse was tested at about 9 months of age using age-sensitive tests of immune status, and then again at about 12 months of age using age-sensitive tests of muscle function. The data were then analyzed using the method of partial least squares to determine the combinations of test weights that maximize the covariance of the weighted sum of immune measures with the weighted sum of muscle function measures. Both AL and CR mice exhibited a statistically significant relation between the immune status tests and the muscle function tests. Maximal covariance was obtained with a set of weighting coefficients consistent with our working hypothesis: mice with high levels of CD4 memory T cells (which increase with age) also had relatively low levels of muscle strength and endurance. Low strength was associated with low CD8 cells in the AL mice, with high numbers of CD8 memory cells in the CR mice and with low CD3 cells in both diet groups. The partial least squares method generates composite indices of immune status and muscle function that can be evaluated as biomarkers of aging rate in these mice. Further work will be needed to assess whether these tests predict either longevity or the trajectory of change in other age-sensitive molecular and physiological traits.