The detection of somatic microsatellite (MS) alterations in tumors is often interpreted as a sign of underlying genomic instability. However, it is unclear why the proportions of altered MS loci vary between different mutator phenotype tumors. We present a simple mathematical analysis that can account for some of these differences, recognizing that the mutations accumulated in a tumor reflect both its mutation rate and number of cell divisions. Only a small proportion of mutated MS loci are expected in tumors with normal or low mutation rates. In contrast, tumors with high mutation rates may or may not acquire mutations depending on the numbers of divisions that proceed the onset of the mutator phenotype. The majority of MS loci should accumulate mutations if high mutation rates are acquired early in tumor progression. Somatic MS mutations provide clues to both the mode and tempo of tumorigenesis.