Eusociality has been recognized as a strong driver of lifespan evolution. While queens show extraordinary lifespans of 20 years and more, worker lifespan is short and variable. A recent comparative study found that in eusocial species with larger average colony sizes the disparities in the lifespans of the queen and the worker are also greater, which suggests that lifespan might be an evolved trait. Here, we tested whether the same pattern holds during colony establishment: as colonies grow larger, worker lifespan should decrease. We studied the mortality of lab-reared Lasius niger workers from colonies at two different developmental stages (small and intermediate-sized) in a common garden experiment. Workers were kept in artificial cohorts that differed only with respect to the stage of the colony they were born in. We found that the stage of the birth colony affected the body size and the survival probability of the workers. The workers that had emerged from early stage colonies were smaller and had lower mortality during the first 400 days of their life than the workers born in colonies at a later stage. Our results suggest that early stage colonies produce small workers with an increased survival probability. These workers are gradually augmented by larger workers with a decreased survival probability that serve as a redundant workforce with easily replaceable individuals. We doubt that the observed differences in lifespan are driven by differences in body size. Rather, we suspect that physiological mechanisms are the basis for the observed differences in lifespan.