Sex ratios are highly variable among parasitic wasps, and the adaptive significance of this variability is currently the subject of intensive study. A model is presented which combines two important effects upon sex ratio selection, (1) local mate competition and (2) host quality. It is proposed that these two effects occur commonly in parasitic wasp species. For species in which females benefit more (in fitness) than do males from large (or good) hosts, the model predicts: 1) Wasps should produce a greater proportion of sons on small (or poor) hosts than on large (or good) hosts. 2) The overall sex ratio should be either greater than or equal to the Hamiltonian sex ratio (i.e. the sex ratio dictated by local mate competition and inbreeding). 3) The overall sex ratio is characterized by three regions, a Hamiltonian sex ratio when few small (or poor) hosts are parasitized, a linearly increasing sex ratio when an intermediate proportion of smaller (or poor) hosts are parasitized, and then a declining sex ratio when many small (or poor) hosts are parasitized. 4) The degree of deviation from the Hamiltonian sex ratio depends on the relative fitness of daughters on small hosts. The less fit are small daughters, the greater an expected sex ratio increase at intermediate frequencies of smaller host parasitization. 5) When a single host size is parasitized, the Hamiltonian sex ratio should be produced. The relevance of the model to various parasitic wasp life histories is discussed.