Telomeres, the short repetitive DNA sequences that cap the ends of linear chromosomes, shorten during cell division and are implicated in senescence in most species. Telomerase can rebuild telomeres but is repressed in many mammals that exhibit replicative senescence, presumably as a tumor suppression mechanism. It is therefore important that we have an accurate understanding of the co-evolution of telomere biology and life-history traits that has shaped the diversity of senescence patterns across species. Gomes et al. (2011) produced a large data set on telomere length (TL), telomerase activity, body mass and lifespan among 57 mammal species. We re-analyzed their data using the same phylogenetic multiple regressions and with several additional analyses to test the robustness of findings. We found substantial inconsistencies in our results compared to Gomes et al.’s. Consistent with Gomes et al. we found an inverse association between TL and lifespan. Contrary to the analyses in Gomes et al., we found a generally robust inverse association between TL and mass, and only weak non-robust evidence for an association between telomerase activity and mass. These results suggest that shorter TL may have been selected for in larger and longer-lived species–likely as a mechanism to suppress cancer. We support this hypothesis by showing that longer telomeres predict higher cancer risk across 22 species. Furthermore, we find that domesticated species have longer telomeres. Our results call into question past interpretations of the co-evolution of telomere biology and life-history traits and stress the need for careful attention to model construction.