Immunosenescence, the age-related decline in immune response, is a well-known consequence of aging. To date, most studies of age-related changes in immune response focused on the cellular and physiological bases of this decline; we have virtually no understanding of the genetic basis of age-related changes in the immune system or if indeed such control exists. We used 25 chromosome substitution lines of Drosophila melanogaster derived from a natural population to address three questions: (i) How is the function of the innate immune system influenced by age? (ii) Is there a genetic basis for phenotypic variation in immune response at different ages? (iii) Is there a genetic basis for differences in the way that age influences the immune function? Virgin females from each line were assayed for immune response using clearance of infection with Escherichia coli at 1 and 4 weeks of age. We found significant genetic variation among lines in immune response at each age. Unexpectedly, when averaged across all lines, the immune response actually improved with age. However, there was significant variation in the effect of age on immune response with 11 lines showing improvement, nine lines showing no change and five exhibiting a decline with age. There was no genetic correlation of immune response across ages suggesting that different loci contribute to variation in immune response at each age. The genetic component of the variation in immune response increased with age, a pattern predicted by the mutation accumulation model of senescence. However, this increase in variation resulted in part from the improvement of the immune response in some lines with age. Thus the observed changes in genetic variation in immune function with age are not entirely explained by the mutation accumulation model.