Sex ratio studies in social Hymenoptera have yielded invaluable tests of kin selection and sex ratio theory. However, relatively few studies exist in facultatively polygynous ant species in which alternative reproductive tactics (independent colony founding versus re-adoption of daughter queens and colony budding) make sex ratio predictions more complex. Generally, a decreased proportional investment in female sexual offspring with increasing degree of polygyny is expected from sex ratio theory. The kin conflict over caste determination hypothesis predicts a larger fraction of females will develop into queens when queen/worker dimorphism is low, and thus it predicts an increased proportional investment in female sexual offspring in microgynous (small queen) colonies. As microgyny and polygyny are correlated in the queen-size-dimorphic ant Leptothorax rugatulus, sex ratio theory and the kin conflict over caste determination hypothesis yield contrasting predictions about its sex ratio patterns. While the former predicts a male bias in the offspring from polygynous, microgynous colonies because relatedness asymmetry is low, the latter hypothesis predicts more new queens due to the lower queen/worker dimorphism that facilitates self-determination of the female larvae. In six populations with different social structure and queen morphology we investigated whether the sex ratio variation in this species accords with the prediction of either hypothesis. At the population level, relative female investment decreased significantly with an increasing degree of polygyny, in accordance with sex ratio theory. At the colony level, sex ratios were highly variable but were correlated neither with polygyny nor queen morphology. Resource level played a significant role in only one population. Furthermore, we found no significant difference in intracolonial relatedness between male- and female-specializing colonies. Thus, individual colony sex ratios do not conform well with current sex ratio theory. The tendency of L. rugatulus microgynes to produce more new queens than macrogynes under laboratory conditions is not reflected in our field data. Consequently, we conclude that microgyny does not constitute an effective selfish strategy of caste disguise in the context of kin conflict over caste determination.