Narrow zones of hybridization between closely related taxa provide valuable natural experiments for examining the evolution of intrinsic barriers to gene exchange. Evolutionary biologists disagree about whether such zones are stable or whether they are transitory, disappearing with the perfection of premating reproductive isolation, with the fusion of the two taxa, or with the extinction of one or the other (e.g., see Mayr, 1963; Bigelow, 1965; Remington, 1968; Paterson, 1978; Templeton, 1981). There is equal uncertainty about the origin of zones. Do they represent secondary contact between forms that have differentiated in allopatry (Mayr, 1963) or are they, in fact, not really hybrid zones but rather the result of differentiation within a continuous series of populations, i.e., parapatric differentiation (Endler, 1977)? Numerous examples of zones have been examined in the past decade using electrophoretic and/or chromosomal differences as genetic markers to identify parental types, F1 hybrids, back-cross individuals, etc. (e.g., Hall and Selander, 1973; Hunt and Selander, 1973; Avise and Smith, 1974; Nevo and Bar-El, 1976; McDonnell et al., 1977; Moran, 1979; Patton et al., 1979; Sage and Selander, 1979; Gartside, 1980; Moran et al., 1980; Harrison and Arnold, 1982). By documenting the distribution of genotypes within and adjacent to the zone it is possible to evaluate the nature and extent of hybridization between the interacting taxa (Patton et al., 1979). In cases where hybridization appears to be limited (where there is an excess of parental types), it is of particular interest to identify barriers to gene exchange that are important in keeping the two taxa distinct. Clearly a first step in characterizing barriers to gene exchange is to determine the extent of post-mating isolation. This paper presents the results of laboratory hybridization between two closely related insect species that are found in an extensive zone of hybridization/overlap in the eastern United States. The data reveal an intriguing asymmetry in the outcome of the two reciprocal interspecific crosses. The uniformity of this interaction over a large distance suggests that secondary contact is the most likely origin of the zone. The field crickets Gryllus pennsylvanicus and G. firmus hybridize and/or overlap in a zone that stretches from the Blue Ridge Mountains in Virginia to southern Connecticut (Fig. 1). Mixed or intermediate populations (determined by electromorph frequencies at an Esterase locus and by distributions of morphological characters) have been found (1) along the Blue Ridge Mountains and in the Shenandoah Valley in Virginia, (2) at the northern end of the Blue Ridge in Maryland, (3) in western New Jersey, (4) at the southern end of the Hudson Valley in New York, and (5) in southern Connecticut (Harrison and Arnold, 1982; see Fig. 1 for localities). The extent of hybridization appears to vary geographically. In Virginia and Maryland, mixed populations of the two eggoverwintering, fall adult Gryllus species consist mostly of crickets that can be classified as one of the two parental types. Presumably, little gene exchange has occurred. In contrast, examination of electrophoretic and morphological characters in mixed/intermediate populations from