The parasitic wasp Nasonia vitripennis has been used extensively in sex allocation research. Although laboratory experiments have largely confirmed predictions of local mate competition (LMC) theory, the underlying assumptions of LMC models have hardly been explored in nature. We genotyped over 3500 individuals from two distant locations (in the Netherlands and Germany) at four polymorphic microsatellite loci to validate key assumptions of LMC theory, in terms of both the original models and more recent extensions to them. We estimated the number of females contributing eggs to patches of hosts and the clutch sizes as well as sex ratios produced by individual foundresses. In addition, we evaluated the level of inbreeding and population differentiation. Foundress numbers ranged from 1 to 7 (average 3.0 +/- 0.46 SE). Foundresses were randomly distributed across the patches and across hosts within patches, with few parasitizing more than one patch. Of the hosts, 40% were parasitized by more than one foundress. Clutch sizes of individual foundresses (average 9.99 +/- 0.51 SE) varied considerably between hosts. The time period during which offspring continued to emerge from a patch or host correlated strongly with foundress number, indicating that sequential rather than simultaneous parasitism is the more common. Genetic differentiation at the regional level between Germany and the Netherlands, as estimated by Slatkin's private allele method (0.11) and Hedrick's corrected G'(LT) (0.23), indicates significant substructuring between regions. The level of population inbreeding for the two localities (F(IL) = 0.168) fitted the expectation based on the average foundress number per patch.