Maternally inherited bacterial endosymbionts can kill male embryos of their arthropod hosts to enhance the transmission efficiency of the endosymbionts. The resources from killed male eggs can be reallocated to infected female hatchlings as additional maternal investment. As a result, the number of offspring per patch and the maternal investment per offspring are expected to differ from the original optimal values for the host mother. Thus, in response to infection, these trait values should be adjusted to maximize the lifetime reproductive success of host females and the fitness of inherited endosymbionts as well. Here, we examined clutch size, egg size, and the proportion of trophic eggs (i.e., production of unhatched eggs, a maternal phenotype) per clutch of host mothers infected with male-killing bacteria. First, we developed a mathematical model to predict the optimal clutch size and trophic egg proportion in uninfected and infected females. Next, we experimentally compared these life-history traits in a ladybird, Harmonia yedoensis, between females infected or uninfected with male-killing Spiroplasma bacteria. Consistent with our predictions, clutch size was larger, egg size was smaller, and trophic egg proportion was lower in infected H. yedoensis females, compared with uninfected females. To our knowledge, this is the first empirical demonstration of variation in these life-history traits depending on infection with bacterial endosymbionts.