The area of sex-ratio theory known as 'local mate competition' is analyzed and extended. New models are presented by adding several realistic assumptions to previous work, including (i) different clutch sizes among females, and (ii) different mating-group sizes within a population. Also, following Hamilton's general formulation for local mate competition, insights into several related sex-ratio phenomena are discussed. These include conflict situations over the sex ratio, and the effect of asymmetric relatedness of parents to sons versus parents to daughters on the sex ratio.The new theory generated in this thesis is applied to data I collected on the sex ratios of Florida fig wasps, Pegoscapus assuetus Grandi and P. jimenezi Grandi, which pollinate the fig species Ficus citrifolia P. Miller and F. aurea Nuttall, respectively. These empirical studies were conducted in the Everglades National Park and on Key Largo, Florida. Due to their unique natural history, fig wasps are an ideal system for the study of local mate competition in natural populations. The number of female fig wasps colonizing an isolated, locally mating patch (= a fig) was experimentally controlled, and the sex ratio as a function of foundress number was measured. The qualitative agreement between predictions and observations is excellent; however, the observed sex ratios are consistently more female biased than predicted. A theoretical exploration of some factors that may explain this quantitative discrepancy is presented. Two a priori candidates are the most compelling at this time: (i) different clutch sizes among females within an isolated, locally mating group, and (ii) genetic differentiation among subpopulations of fig wasps. Empirical methods for the further study of fig wasp sex ratios are discussed extensively. Additionally, information and discussion is presented on the natural history of Florida fig wasps, their nematode and wasp parasites, the mechanism of sex-ratio adjustment, and breeding structure of the wasps, including subpopulation structure, the number of foundresses per fig and the cues assessed by the wasps when choosing a fig in which to oviposit.Fig wasps provide an excellent opportunity for evaluating sex-ratio theory and the success of population-genetic models in predicting behavior patterns accurately. The wasps are easily observed, important parameters can be readily quantified in natural populations, and they are amenable to experimentation in the field. Also, since there are 900 fig wasp species, this group provides a rich source of comparative information.