Simple SummaryAnimals engage in various symbioses. However, these interactions are not always beneficial for the host; they can also incur costs under certain circumstances. The bacterial symbiont supports, on the one hand, the cuticle formation of the sawtoothed grain beetle Oryzaephilus surinamensis, which is extremely beneficial under dry conditions as a thicker and more melanized cuticle prevents desiccation of the insect. On the other hand, under higher humidity, the benefit is strongly reduced. In this study, we investigated whether harboring a symbiont can also be a disadvantage. Therefore, we first measured the number of symbionts throughout the beetles’ life and found a strong increase during the end of metamorphosis, just before beetles reach adulthood. Afterwards, males lose the symbionts again, whereas females retain a stable number. A comparison of beetles with and without symbionts revealed no differences in many life history traits. Larval development took the same time and there was also no difference in adult mortality or lifespan or the number of offspring of females. However, females with symbionts started to reproduce significantly later by one to two weeks, meaning they have a disadvantage in comparison to females without symbionts. Thus, harboring a symbiont is beneficial or costly in a context-dependent manner.Animals engage in a plethora of mutualistic interactions with microorganisms that can confer various benefits to their host but can also incur context-dependent costs. The sawtoothed grain beetle Oryzaephilus surinamensis harbors nutritional, intracellular Bacteroidetes bacteria that supplement precursors for the cuticle synthesis and thereby enhance desiccation resistance of its host. Experimental elimination of the symbiont impairs cuticle formation and reduces fitness under desiccation stress but does not disrupt the host’s life cycle. For this study, we first demonstrated that symbiont populations showed the strongest growth at the end of metamorphosis and then declined continuously in males, but not in females. The symbiont loss neither impacted the development time until adulthood nor adult mortality or lifespan. Furthermore, lifetime reproduction was not influenced by the symbiont presence. However, symbiotic females started to reproduce almost two weeks later than aposymbiotic ones. Thus, symbiont presence incurs a metabolic and context-dependent fitness cost to females, probably due to a nutrient allocation trade-off between symbiont growth and sexual maturation. The O. surinamensis symbiosis thereby represents an experimentally amenable system to study eco-evolutionary dynamics under variable selection pressures.